Influence of austenitizing parameters on microstructure and mechanical properties of Al-Si coated press hardened steel

Influence of austenitizing parameters on microstructure and mechanical properties of Al-Si coated press hardened steel

Materials and Design 172 (2019) 107707 Contents lists available at ScienceDirect Materials and Design journal homepage: www.elsevier.com/locate/matd...

4MB Sizes 0 Downloads 69 Views

Materials and Design 172 (2019) 107707

Contents lists available at ScienceDirect

Materials and Design journal homepage: www.elsevier.com/locate/matdes

Influence of austenitizing parameters on microstructure and mechanical properties of Al-Si coated press hardened steel Lindsay Golem 1, Lawrence Cho 2, John G. Speer, Kip O. Findley ⁎ Advanced Steel Processing and Products Research Center, Colorado School of Mines, Golden, CO 80401, USA

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Smaller PAGS is generally beneficial for mechanical performance of Al-Si coated PHS. • Tensile fracture/ductility is independent of austenitizing conditions of these PHS. • Quasi-cleavage occurs under plane strain in PHS austenitized at high temperatures. • Bendability/toughness decreases for AlSi coated PHS austenitized at high temperatures. • Both Al-Si coating and martensite microstructure affect fracture behavior.

a r t i c l e

i n f o

Article history: Received 10 August 2018 Received in revised form 13 March 2019 Accepted 15 March 2019 Available online 16 March 2019 Keywords: Press hardened steel Hot stamping Prior austenite grain size Al-Si coating Martensite Mechanical behavior

a b s t r a c t The influence of the austenitizing parameters on the microstructure and mechanical behavior of Al-Si coated press hardened 22MnB5 steel was evaluated in this study. Increasing the austenitizing temperature and hold time resulted in microstructural homogenization, an increase in the prior austenite grain size (PAGS), and increased thickness of the substrate-coating interdiffusion layer. The mechanical properties were evaluated using smooth-sided tensile testing, double edge-notch tensile testing, and free bend testing. Refinement of the PAGS was, in general, beneficial for the mechanical performance of the press hardened steel (PHS); i.e. strength, notch displacement, maximum bending load, and bend angle at maximum load all generally increased with decreasing PAGS. The stress and strain state experienced during each type of test influenced the sensitivity of mechanical and fracture behavior of the PHS to austenitizing conditions. Furthermore, the thickening of the substrate-coating interdiffusion layer, composed of brittle intermetallic phases, at higher austenitizing temperature negatively influenced bendability of the PHS. Ductile fracture was commonly observed in the steel matrix composed of a lath martensitic structure. However, quasi-cleavage fracture was observed for the martensitic matrix with the largest PAGS, which implies a loss of toughness, possibly because of the increased substructure size. © 2019 Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).

1. Introduction ⁎ Corresponding author. E-mail address: kfi[email protected] (K.O. Findley). 1 Current address: Gestamp North American R&D Center, Auburn Hills, Michigan 48326, USA. 2 Current address: National Institute of Standards and Technology, Boulder, CO 80305, USA.

Vehicle lightweighting efforts have remained a high priority in recent years due to the increase in the Corporate Average Fuel Economy standard. A variety of advanced high strength steels (AHSS) have been identified as viable options for achieving lighter vehicle architectures

https://doi.org/10.1016/j.matdes.2019.107707 0264-1275/© 2019 Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

2

L. Golem et al. / Materials and Design 172 (2019) 107707

through increased strength. Of these, press hardened martensitic steels have increasingly been implemented into structural components, e.g. anti-intrusion applications such as B-pillars, because of their ultrahigh specific strength, high anti-intrusion properties, absence of spring back, and adequate plasticity. However, the resulting martensitic microstructure, which is in the as-quenched or auto-tempered condition, has lower ductility than other classes of AHSS, leading to concerns about crashworthiness, which is a critical metric for structural components. “Toughness” is considered to be a metric relevant to crashworthiness. Laboratory evaluation of crashworthiness has proven to be difficult using conventional fracture toughness testing methods due to the non-plane strain conditions present in thin sheet materials [1]. Therefore, methods such as stacked Charpy testing [2], bend testing [3,4], and notched tensile testing [3,4] have been used to measure toughness of a press hardened sheet steel with respect to resistance to crack initiation and propagation. Previous studies have shown that the mechanical and fracture behavior of press hardened steel (PHS) can be influenced by microstructural features such as prior austenite grain size (PAGS) [2,5–7], surface microstructure and properties [8], microstructure homogeneity [9], and precipitates [10]. A decrease of austenitizing temperature in the hot stamping process results in the refinement of the PAGS of a PHS, leading to the improvement of the tensile properties and impact toughness [2,5–7,9]. The PAGS-dependent mechanical properties of PHSs have been investigated most frequently by means of uniaxial tensile testing [5–7] and in less frequent cases, Charpy impact testing [9], and free bend testing. However, direct correlation of properties generated from these experiments in the context of the stress states associated with each test has not been performed. In addition, there is limited information in literature on the impact of the coating microstructure and properties on the room temperature mechanical properties of PHS. Instead, investigations on Al-Si coated PHSs have focused on the wear behavior [11], high-temperature mechanical behavior [12], corrosion resistance [13], and cracking and interfacial debonding of the coating on a PHS [14,15]. Most importantly, there is a need to investigate the complicated relationship between mechanical properties and interrelated microstructural factors in the substrate and coating, which can simultaneously change through differences in austenitizing parameters during hot stamping. The present study focused on understanding the effect of austenitizing parameters (austenitizing temperature and hold time) on the mechanical and fracture behavior of Al-Si coated press hardened 22MnB5 steel. The study also aimed to reveal the correlation between the various microstructural features, such as PAGS and coating microstructure and thickness, and the mechanical properties of Al-Si coated PHS. Multiple test methods, i.e. smooth-sided tensile testing, double edge-notch tensile (DENT) testing, and free bend testing, were employed to evaluate the room temperature mechanical properties, particularly toughness, of the Al-Si coated PHS with different applied stress and strain states. The ability of the different test methods to detect differences in mechanical behavior between the various austenitizing parameters was evaluated. The evaluated mechanical properties included strength, ductility, notch sensitivity, notch displacement, and bend angle at maximum load. 2. Experimental A cold-rolled, full hard Al-Si coated 22MnB5 steel was used in the present study. The thickness of the cold-rolled sheet was 1.5 mm. The microstructure of the steel in the cold-rolled condition consisted of ferrite and pearlite. The chemical composition is given in Table 1. All material used for this study was hot-dip Al-Si coated prior to heat treatment in order to prevent decarburization and oxidation of the underlying Table 1 Chemical composition of the 22MnB5 alloy (in wt%). C

Mn

Si

Cr

Ti

Al

S

P

B

0.24

1.16

0.21

0.22

0.031

0.042

0.002

0.004

0.003

material during the austenitization stage in the hot stamping process. The coating applied to this material consisted of approximately 85 wt % Al, 10 wt% Si, and 5 wt% Fe with an average coating thickness of 25 μm per side prior to hot stamping. The Ac1 and Ac3 temperatures of the steel, measured at a heating rate of 28 °C/h by means of dilatometry, were 714 °C and 863 °C, respectively. The hot stamping was performed using the Gestamp® HardTech laboratory-scale hot stamping line. Al-Si coated steel blanks were first heated to an austenitizing temperature in the range of 850–1025 °C and isothermally held for 3–30 min in a small roller hearth furnace. The steel blanks were then transferred to water-cooled flat dies where the blanks were subsequently quenched to room temperature without imposed deformation. Overall, the austenitizing condition matrix was established to evaluate a range of PAGS as well as assess possible steel microstructure inhomogeneity and coating microstructure effects. An austenitizing temperature of 850 °C was selected to measure the influence of potential chemical and microstructural inhomogeneities. Additionally, a range of hold times at 850 °C was investigated to potentially allow more time for homogenization. Austenitizing temperatures of 900 °C and 930 °C were chosen to provide processing conditions consistent with the standard range used industrially for hot stamping. The upper limit temperature, 1025 °C, was selected to maximize the range of PAGS evaluated. Microstructural analysis was performed on the through-thickness plane with the rolling direction (RD) for each specimen perpendicular to the analysis plane. The cross-sections of the specimens cut from the heat-treated blanks were mounted, mechanically ground, and polished. A final polishing step was performed using a vibratory polisher and colloidal silica with a particle size of 0.06 μm. Etching was conducted using two different methods: a 2% nital solution to observe the substructure of the martensitic matrix and the microstructure of the substrate-coating interdiffusion layer, and a saturated aqueous picric acid etchant with Teepol® surfactant heated to 65 °C in order to reveal the prior austenite grain boundaries. The microstructure of the steel matrix and substrate-coating interdiffusion layer of the heat-treated specimens was analyzed by means of light optical microscopy and field emission scanning electron microscopy (SEM) equipped with energy dispersive spectroscopy (EDS) and electron back scatter diffraction (EBSD). The SEM observation was conducted in a JEOL 7000F FE-SEM. The average PAGS for each specimen was calculated using the concentric circle method described in ASTM E112 [16]. The specimens austenitized at 850 °C, which is slightly below the Ac3 temperature, contained small amounts of untransformed ferrite in the asquenched microstructure. The ferrite area fraction, which is equivalent to the ferrite volume fraction, of each specimen was measured by manually selecting the ferrite grain perimeter using Image Pro Plus® version 6.0 image analysis software. Light optical microscopy was used to measure the thickness of the interdiffusion layer, and ten measurements were performed for each specimen to determine the average. Mechanical behavior was evaluated using smooth-sided tensile testing, DENT testing, and free bend testing. For smooth-sided (unnotched) tensile testing, ASTM E8 tensile specimens with a gauge length of 50 mm were prepared by abrasive water-jet machining. A schematic of the specimen geometry for smooth-sided tensile testing is shown in Fig. 1(a). All specimens were tested at a crosshead speed of 0.042 mm/s until failure, and displacement was measured using a 50.8 mm extensometer centered on the reduced cross section. Loaddisplacement data were then used to calculate the yield strength, ultimate tensile strength, strain to failure, uniform strain, and post uniform strain. For DENT testing, specimens were electrical discharge machined parallel to the RD. On either side of the DENT specimen, notches with a depth of 2.5 mm and radius of 0.2 mm were machined to allow symmetrical loading of the specimen along the centerline (Fig. 1(b)). All specimens were displaced at a rate of 0.042 mm/s and pulled to fracture. DENT testing incorporated the use of 2-D digital image correlation (DIC) to measure localized displacements around the notched regions. Notch

L. Golem et al. / Materials and Design 172 (2019) 107707

3

12.7 mm

20 mm

20 mm/min

50 mm

(a) ND

2.5 mm 25.4 mm

12.7 mm

R=0.20 mm

TD

RD

50 mm

45°

(b)

R=0.4 mm 2.5 mm

Notch tip Notch displacement

30.93 mm

3.35 mm

(c)

(d)

Fig. 1. (a) Specimen geometry for smooth-sided tensile testing in accordance with the ASTM E8 standard. (b) Specimen geometry for DENT testing. (c) DIC strain contour map of a DENT specimen generated in ARAMIS® DIC software showing the analysis location for notch displacement evaluation. (d) Testing set up for free bend testing in accordance with the VDA238-100 standard [18].

displacement, i.e. displacement between two fixed points above and below the notch opening, was measured by means of DIC as described in a recent study by Golem [17]. Fig. 1(c) shows a representative strain contour map indicating the location for notch displacement assessment. Each point was located approximately 0.6 mm inwards from the edge of the specimen. Free bend testing was conducted using a MTS hydraulic frame with tooling prepared in accordance with the VDA238-100 standard [18]. Fig. 1(d) shows the test configuration used for free bend testing. The specimen dimensions were 60 mm × 30 mm × 1.5 mm with the longer side of the specimen parallel to the transverse direction. Testing was conducted with a 20 mm/min punch speed. The punch radius was 0.4 mm. The fracture surfaces of the specimens after all tests were examined in a JEOL 7000F FE-SEM. For a selected condition, the fracture surface of the DENT specimen was analyzed using 3D digital optical microscope (Keyence VHX-5000).

due to the short hold time (3 min). As the hold time extended to 30 min (Fig. 3(b)), ferrite grains were still present along with martensite. As the temperature increased to or above 900 °C, the microstructure was fully martensitic and did not contain any ferrite grains. Fig. 3(c) shows the microstructure of the PHS austenitized at 900 °C, which is above the Ac3 temperature, indicating the absence of ferrite grains. Additionally, the coarse precipitates were not observed in the martensitic matrix of the steels austenitized at 850 °C for 30 min (Fig. 3(b)) or 900 °C for 6.5 min (Fig. 3(c)). It should be pointed out that the microstructure of the PHS in the as-quenched state may contain a small fraction of inter-lath retained austenite, which could not be identified in the present study.

Ac3 = 863 °C 40

3.1. Microstructure The influence of austenitizing parameters on the PAGS of the PHS is shown in Fig. 2 (see also Supplementary Fig. S1). The PHS were austenitized in the range of 850–1025 °C and held isothermally for 3–30 min prior to die-quenching. The PAGS of the PHS increased with increasing austenitizing temperature. At a given austenitizing temperature, an increase in the hold time led to an increase in the measured PAGS. The total range of PAGS was approximately 30 μm, but the range observed between hold times at a given austenitizing temperature was only 2–6 μm, confirming that the austenitizing temperature has a more significant influence on the change in PAGS than the hold time. Figs. 3(a) and (b) show SEM micrographs of the substrate microstructure of the PHS austenitized at 850 °C, which was slightly below the Ac3 temperature (863 °C). When the steel was austenitized at 850 °C for 3 min (Fig. 3(a)), the microstructure consisted of a lath martensitic matrix with very small amounts of untransformed ferrite (b1 vol%). The ferrite grains are characterized by a polygonal morphology and a deep etching response as compared to the surrounding lath martensitic matrix. Coarse precipitates with a diameter in the range of 150–200 nm were occasionally observed in addition to lath martensite, likely undissolved cementite

Prior Austenite Grain Size, m

3. Results

30

20

10

0 840

t = 3 min t = 6.5 min t = 10 min t = 30 min

880 920 960 1000 Austenitizing Temperature, °C

1040

Fig. 2. PAGS as a function of austenitizing temperature and hold time (t) for the PHS. Error bars represent the standard deviation in the measured PAGS for each condition.

4

L. Golem et al. / Materials and Design 172 (2019) 107707

850 °C-3 min

850 °C-30 min

Ferrite

Ferrite 1μm

1μm

(a)

(b) 900 °C-6.5 min

1μm

(c) Fig. 3. SEM micrographs of the microstructure of the PHS austenitized at (a) 850 °C for 3 min, (b) 850 °C for 30 min, and (c) 900 °C for 6.5 min. The black arrows in (a) indicate precipitates, likely carbides that were not dissolved during austenitizing.

3.2. Mechanical properties Smooth-sided tensile testing was conducted for each austenitizing condition to evaluate strength and ductility of the Al-Si coated PHS. Fig. 5 summarizes the smooth-sided tensile test results for the Al-Si

80 Interdiffusion Layer Thickness, m

The retained austenite volume fraction of boron-alloyed steels in the asquenched state is less than 2 vol% [19,20]. Prior to hot stamping, the coating of hot-dip Al-Si coated 22MnB5 steel consists of the fcc-Al phase, as a main constituent, and various Fe-Al-Si intermetallic compounds. The Al-Si coating transforms into a complex microstructure comprised of various Fe-Al or Fe-Al-Si intermetallic compounds during the austenitizing heat treatment in the hot stamping process [14,21]. The phases or intermetallic compounds observed in the coating after hot stamping are a function of the austenitizing conditions, which affect the diffusion kinetics of Fe, Al, and Si [14,21]. The influence of the austenitizing conditions on the thickness of the substrate-coating interdiffusion layer of the Al-Si coated PHS is shown in Fig. 4 (see also Supplementary Fig. S2). Increasing the temperature and/or hold time results in more interdiffusion and reaction of Fe, Al, and Si, which leads to a change of the major constituent in the coating from Fe2Al5 to FeAl and an increase in the thickness of the interdiffusion layer. In the case of the specimen austenitized at 1025 °C for 30 min, the average interdiffusion layer thickness was as great as 76 μm, more than 5% of the total thickness of the sheet. The microstructural features, including the PAGS, ferrite volume fraction, and interdiffusion layer thickness, quantified in the present work for the Al-Si coated PHS are summarized in Table 2.

t = 3 min t = 6.5 min t = 10 min t = 30 min

60

40 Initial coating thickness 20 840

880 920 960 1000 Austenitizing Temperature, °C

1040

Fig. 4. Average thickness of substrate-coating interdiffusion layer as a function of austenitizing temperature and hold time (t). Error bars represent the standard deviation in the measured interdiffusion layer thickness for each condition.

L. Golem et al. / Materials and Design 172 (2019) 107707

5

Table 2 PAGS, ferrite volume fraction, and substrate-coating interdiffusion layer thickness of the Al-Si coated PHS after austenitizing. Austenitizing temperature (°C)

Hold time (min)

Avg. PAGS (μm)

Ferrite fraction (vol%)

Avg. interdiffusion layer thickness (μm)

850 850 850 900 930 930 1025 1025

3 6.5 30 6.5 6.5 30 10 30

5.7 ± 0.42 6.8 ± 0.33 7.1 ± 1.25 11.8 ± 0.58 16 ± 3.80 22 ± 12.40 32.3 ± 1.96 34.5 ± 2.83

0.16 0.13 0.46 – – – – –

27.0 ± 3.13 32.4 ± 4.89 43.3 ± 2.05 37.0 ± 4.94 41.5 ± 4.10 51.9 ± 2.76 58.1 ± 2.08 75.6 ± 1.88

1180 t = 3 min t = 6.5 min t = 10 min t = 30 min

1160 1140 1120 1100 1080 840

880

920

960

uniform elongation showed no significant difference as a function of the austenitizing conditions. The average total elongation ranged from 7.5 to 7.8%. The average post uniform elongation ranged from 1.9 to 2.2%. The small fraction of ferrite present at the lowest austenitizing temperature did not appear to have a substantial influence on the strength. DENT testing was conducted to evaluate notch sensitivity and fracture behavior in the presence of a stress concentration, as a function of austenitizing parameters. Fig. 6 shows the DENT test results for the AlSi coated PHS austenitized under different conditions (see also Supplementary Fig. S3(b)). Notch tensile strength decreased with increasing austenitizing temperature or hold time (Fig. 6(a)), indicating a clear PAGS-dependence of the notch tensile strength.

Ultimate Tensile Strength, MPa

0.2% Offset Yield Strength, MPa

coated PHS austenitized under different conditions. Engineering stress versus engineering strain curves for three of the austenitizing conditions are provided in Supplementary Fig. S3(a). All conditions showed continuous yielding behavior and some post uniform ductility. The work hardening behavior was not significantly different between the different austenitizing conditions. The primary difference between each austenitizing condition was with respect to strength. Both yield strength and ultimate tensile strength of the Al-Si coated PHS generally decreased with increasing austenitizing temperature or hold time (Figs. 5(a) and (b)). The total elongation, often used as an indication of the material ductility, and post uniform elongation were plotted as a function of austenitizing parameters in Figs. 5(c) and (d). Total elongation and post

1000 1040

1580 1560 1540 1520 1500 840

880

920

960

1000 1040

Austenitizing Temperature, °C

Austenitizing Temperature, °C

(a)

(b)

8.25

2.50

Post Uniform Elongation, %

Total Elongation, %

1600

8.00 7.75 7.50 7.25 7.00 840

880

920

960

1000 1040

2.25

2.00

1.75

1.50 840

880

920

960

1000 1040

Austenitizing Temperature, °C

Austenitizing Temperature, °C

(c)

(d)

Fig. 5. Summary of the smooth-sided tensile test results for the Al-Si coated PHS. Plots of (a) 0.2% offset yield strength, (b) ultimate tensile strength, (c) total elongation, and (d) post uniform elongation, as a function of austenitizing temperature and hold time (t). Error bars represent the standard deviation in the measured values for three replicate tests.

6

L. Golem et al. / Materials and Design 172 (2019) 107707

Notch displacement to fracture was employed as a possible measure of fracture resistance. This metric is analogous to crack tip opening displacement, which is a fracture mechanics parameter associated with the strain field and stress intensity ahead of a crack tip. Notch displacement to fracture may also be considered as a measure of local fracture ductility in the notched specimen. Notch displacement decreased slightly with increasing austenitizing temperature or hold time (Fig. 6(b)), in contrast to the total elongation results obtained from smooth-sided tensile tests (Fig. 5(c)). In particular, the decreased notch displacement for the 1025 °C-30 min condition, which is clearly lower than all other conditions, may be due to the coarse prior austenite grain structure, as will be discussed in Section 4.2. Fig. 7 shows the free bend test results for the Al-Si coated PHS (see also Supplementary Fig. S3(c)). Fig. 7(a) plots the maximum load during the free bend test as a function of austenitizing parameters. The maximum load generally decreased with increasing austenitizing temperature and hold time, which was associated with increased PAGS and interdiffusion layer thickness. At the same time, the bend angle at maximum load, which has commonly been used as another metric for evaluating toughness and crash resistance of a material [22], generally decreased with increasing austenitizing temperature and hold time (Fig. 7(b)). There is, however, a data set that falls slightly outside of the general trend, which is the bend angle for the specimens austenitized at 850 °C for the short hold times (3 min and 6.5 min). 3.3. Fractography Three austenitizing conditions (850 °C-3 min, 900 °C-6.5 min, and 1025 °C-30 min) were selected for fractographic analysis to help understand the critical relationships between process, microstructure, and performance. The specimen austenitized at 850 °C for 3 min represents the lowest austenitizing temperature. The austenitizing condition of 900 °C for 6.5 min is within the standard range of conditions used industrially for hot stamping. The specimen processed at 1025 °C for 30 min was chosen to investigate the influences of large PAGS and interdiffusion layer thickness on the fracture behavior. Fig. 8 shows SEM micrographs of the fracture surfaces of the specimens after smooth-sided tensile tests. The micrographs were recorded at the mid-thickness region of the specimens. The fracture was ductile for all conditions, as can be inferred from the dimpled (microvoid) fracture surface appearance obtained for all three austenitizing conditions: 850 °C-3 min, 900 °C-6.5 min, and 1025 °C-30 min.

Fig. 9(a) shows a representative fracture surface of a DENT specimen at low magnification. For all the austenitizing conditions, the DENT specimen fracture surface shows a distinct triangular flat fracture region ahead of both notch tips. The region outside of the triangle undergoes slant fracture typical of a biaxial or plane stress condition. The slant fracture plane was inclined at approximately 45° to the tensile axis (Fig. 9 (a)). Figs. 9(b) to (d) show SEM micrographs of the fracture surfaces of the three selected specimens after DENT testing. In the case of the specimens austenitized at 850 °C for 3 min and 900 °C for 6.5 min, the flat fracture region exhibited entirely ductile fracture (Figs. 9(b) and (c)). For the specimen austenitized at the highest temperature for the longest time, 1025 °C for 30 min, cleavage steps and tear ridges, indicative of quasi-cleavage fracture, are noted in the flat fracture region ahead of the notch tip (Fig. 9(d)). The slant fracture region exhibited entirely ductile fracture for all the austenitizing conditions examined. In general, the dimple size in the triangular flat fracture region was much larger than in the slant fracture region (Figs. 9(b) and (c)). Fig. 10 shows SEM micrographs of the fracture surfaces of specimens after free bend tests. In the case of free bend testing, the importance of the surface microstructure and coating characteristics is expected to be greater as compared to the other two testing methods. The SEM analysis, therefore, focused on the fracture surface near the substratecoating interdiffusion layer (right column of Fig. 10) as well as the mid-thickness region of the specimen (left column of Fig. 10). The fractographic observation at the mid-thickness region of the specimens austenitized at 850 °C for 3 min and 900 °C for 6.5 min revealed a dimpled or microvoid coalescence fracture surface (Fig. 10(a) and (b)), consistent with the tensile test results. For the specimen austenitized at 1025 °C for 30 min, the fracture surface consisted of a mixture of dimples and quasi-cleavage facets with steps and tear ridges (Fig. 10(c)). Overall, the fractographic analysis conducted near the interdiffusion layer indicated that the interdiffusion layer, consisting mainly of Fe-Al intermetallic phases, is more brittle than the martensitic steel substrate. For all the austenitizing conditions, the fracture surface consisted mainly of flat regions in the upper part of the interdiffusion layer. The fracture path for these flat regions is likely along the grain boundaries of the Fe-Al intermetallic crystals. Iron aluminide (FeAl), which is one of the main constituents in the interdiffusion layer, is known to be sensitive to brittle intergranular fracture [23]. River marks, which consist of cleavage steps or tear ridges, were present in the lower part of the coating area. An intergranular crack along the grain boundary of the Fe-Al intermetallic crystals is clearly shown in Fig. 10(c).

0.24 t = 3 min t = 6.5 min t = 10 min t = 30 min

1800 1760 1720 1680 1640 840

880

920

960

1000 1040

Notch Displacement, mm

Notch Tensile Strength, MPa

1840

0.22

0.20

0.18

0.16 840

880

920

960

1000 1040

Austenitizing Temperature, °C

Austenitizing Temperature, °C

(a)

(b)

Fig. 6. Summary of the DENT test results for the Al-Si coated PHS. Plots of (a) notch tensile strength and (b) notch displacement as a function of austenitizing temperature and hold time (t). Error bars represent the standard deviation in the measured values for three replicate tests.

Bending Angle at Maximum Load, °

L. Golem et al. / Materials and Design 172 (2019) 107707

Maximum Load, N

2700 t = 3 min t = 6.5 min t = 10 min t = 30 min

2600 2500 2400 2300 2200 2100 840

880

920

960

1000 1040

7

60

55

50

45

40 840

880

920

960

1000 1040

AustenitizingTemperature, °C

AustenitizingTemperature, °C

(a)

(b)

Fig. 7. Summary of the free bend test results for the Al-Si coated PHS. Plots of (a) maximum load and (b) bending angle at maximum load as a function of austenitizing temperature and hold time (t). Error bars represent the standard deviation in the measured values for three replicate tests.

indicates the degree of stress triaxiality, measured as the ratio of the mean stress to effective von Mises stress, that evolves in critical locations during each test method. Smooth-sided tensile testing has the least complex loading condition and produces the lowest degree of triaxiality (0.33) with the loading being uniaxial tension up to necking. In DENT testing, the notch creates a triaxial stress state ahead of the notch tip while near the surface or edge of the sample, a biaxial stress state is present

4. Discussion 4.1. Influence of stress and strain states on mechanical and fracture behavior Table 3 compares the specific stress and strain states imposed during each of the three mechanical tests used in the present study. It also

850°C-3 min

10μm

900°C-6.5 min

2μm

10μm

(a)

2μm

(b) 1025°C-30 min

10μm

2μm

(c) Fig. 8. SEM micrographs of the fracture surface of smooth-sided tensile specimens austenitized at (a) 850 °C for 3 min, (b) 900 °C for 6.5 min, and (c) 1025 °C for 30 min.

8

L. Golem et al. / Materials and Design 172 (2019) 107707

2D View 1mm

3D View

Slant Fracture (Biaxial)

Triangular Flat Fracture (Triaxial)

Notch tip

(a) Triangular Flat Fracture Zone

Triangular Flat Fracture Zone

Triangular Flat Fracture Zone

Slant Fracture Zone

Slant Fracture Zone

Slant Fracture Zone

2μm

(b)

2μm

(c)

2μm

(d)

Fig. 9. (a) 2D and 3D optical micrographs of the fracture surface of a DENT specimen austenitized at 900 °C for 6.5 min indicating the flat fracture region ahead of either notch tip as well as the slant fracture region. SEM micrographs of the flat fracture region ahead of a notch tip and the slant fracture region in the specimens austenitized at (b) 850 °C for 3 min, (c) 900 °C for 6.5 min, and (d) 1025 °C for 30 min.

because there is no material constraint creating an opposing stress. Free bend testing differs from the other two test methods in that the maximum stress and strain occur on the outer surface of the material, and the surface region experiences a plane strain tension state during free bend testing [24]. Therefore, the free bend testing heavily emphasizes the influence of the surface microstructure along with the substrate properties. Both DENT testing and free bend testing result in higher degrees of triaxiality in the region of maximum strain compared to smooth-sided tensile testing. Finite element analysis by Pape [25] showed that for a DENT specimen with a notch radius of 0.45 mm, the stress triaxiality was approximately 0.55 up to an equivalent plastic strain of 1.6. Roth and Mohr [24] conducted a finite element simulation of a V-bend test for a dual phase steel and showed that the triaxiality was 0.56 for an equivalent plastic strain up to 0.5. The results of the above-referenced studies indicate that the degree of triaxiality in the region of maximum strain for these two test methods may be comparable.

The results of the present study suggest that the stress and strain state in each test method influence the mechanical behavior and fracture behavior of Al-Si coated PHS. An example is the notch strengthening effect, i.e. the notch tensile strength is higher than the tensile strength obtained from smooth-sided tensile specimens (Figs. 5 (b) and 6(a)). This effect is associated with the stress concentration in notched specimens, which leads to constrained plastic deformation mainly in a small zone ahead of the notch [26]. Notched specimens subjected to a triaxial stress state have a lower maximum shear stress for a given nominal applied stress due to the presence of three stress components, and thus yield at a higher applied (maximum principal) stress than specimens subjected to a uniaxial or biaxial stress state. The triaxial stress state that develops ahead of a notch also contributes to the fracture process, e.g. the triangular flat fracture shown in Fig. 9(a) is associated with a triaxial stress state, more specifically high local triaxiality near the notch tip [27–29]. The triangular morphology

L. Golem et al. / Materials and Design 172 (2019) 107707

Near the Mid-Thickness

9

Near the Coating Surface

850°C-3 min Coating

10μm

α´-martensite

10μm

(a) 900°C-6.5 min

Coating

10μm

α´-martensite

10μm

(b) 1025°C-30 min

Coating

10μm

α´-martensite

10μm

(c) Fig. 10. SEM micrographs of the fracture surface of bending specimens austenitized at (a) 850 °C for 3 min, (b) 900 °C for 6.5 min, and (c) 1025 °C for 30 min.

Table 3 Comparison of difference in stress and strain state between the mechanical testing methods. Degree of stress triaxiality is defined as the ratio of the mean stress to effective von Mises stress.

Stress and strain state Degree of stress triaxiality Remarks

Smooth-sided tensile testing

DENT testing

Free bend testing

Uniaxial tension

Plane strain tension on the outer surface

1/3

Triaxial stress state in center of thickness ahead of notch ~0.55 (notch radius = 0.45 mm) [25]

Measures mechanical characteristics of steel bulk microstructure

Measures mechanical characteristics of steel bulk microstructure

Measures mechanical characteristics of both surface coating and steel bulk microstructure

~0.56 (V-bend test of a dual phase steel) [24]

L. Golem et al. / Materials and Design 172 (2019) 107707

of the flat fracture region results from the transition in stress state during ductile fracture of a sheet material. This type of fracture is more likely to appear in severely notched conditions that produce a high degree of stress triaxiality [27,29]. Asserin-Lebert et al. [27] conducted interrupted tensile tests with notched Al alloy samples and described the failure mechanism. During ductile fracture of a sheet material, the fracture initiates ahead of the notch root, and the flat triangular fracture zone is formed when the maximum load is reached. The high local stress triaxiality in the triangular flat zone promotes void growth and coalescence [27–29], which explains the observation of coarse dimples near the notch root (Figs. 9(b) and (c)). Upon further loading, the stress state near the crack front transitions to a biaxial stress state with crack propagation occurring by tearing, typically on a 45° plane, as explained by Faccoli et al. [28]. A comparison between the different mechanical test results illustrates how the stress and strain state experienced during testing affects the observed sensitivity of mechanical behavior to austenitizing conditions. For instance, the tensile elongation was independent of the austenitizing conditions or PAGS (Fig. 5(c)), whereas the notch displacement and bending angle at maximum load had a clear PAGS-dependence (Figs. 6(b) and 7 (b)). These results suggest that tensile property data alone are not sufficient to predict performance metrics related to different stress and strain states. The importance of stress and strain state is also apparent when the fracture surface appearance after each mechanical test is compared. In the case of DENT testing and bend testing, a transition from dimpled fracture to quasi-cleavage fracture was observed as the austenitizing temperature increased from 900 °C to 1025 °C (Figs. 9 and 10). However, this transition was not observed for the smooth-sided tensile test results. That is, the DENT and bending specimens that experienced higher triaxiality were more sensitive to the activation of brittle, quasi-cleavage fracture mechanisms. The results, therefore, suggest that multiple test methods with various stress and strain states may be needed to correlate laboratory scale testing to toughness or crash performance of sheets used for automotive structural parts. 4.2. Influence of austenitizing parameters on strength and toughness Refinement of the PAGS was, in general, beneficial to the mechanical performance of Al-Si coated PHS. Strength and maximum bending load generally decreased with increasing austenitizing temperature and hold time, i.e. increasing PAGS. In particular, yield strength increased in inverse proportion to the square root of the PAGS, illustrating a HallPetch strengthening trend, as shown in Fig. 11. The slope of the HallPetch plot of 0.2% offset yield strength as a function of PAGS was calculated as 9.3 MPa·mm0.5. This slope represents the locking parameter, ky, in the Hall-Petch equation. The ky value obtained for the Al-Si coated press hardened 22MnB5 steel was relatively low, which indicates less dependence of yield strength on PAGS as compared to values reported for other lath martensitic steels (9.5–17.5 MPa·mm0.5) [30,31]. It is known that the ky is associated with the effectiveness of boundary strengthening and that the ky value for martensitic steels varies with the alloying levels [30,31] and degree of tempering [31]. Fig. 11 also shows that the smooth specimen tensile strength and notch tensile strength increased with decreasing PAGS, following a Hall-Petch type relationship, similar to the smooth specimen yield strength. The ky for the notch tensile strength (14.1 MPa·mm0.5) was greater as compared to the smooth specimen tensile strength (8.7 MPa·mm0.5), which may indicate a greater sensitivity of the strength to PAGS when subjected to a triaxial stress state. The present study provides evidence indicating a negative influence of a coarse prior austenite grain structure on the ductility and toughness of Al-Si coated PHS. Notch displacement, which is related to local ductility to fracture in the presence of the notch, generally decreased with increasing austenitizing temperature and hold time, i.e. increasing PAGS (Fig. 6(b)). In particular, the noticeable decrease in notch displacement at the highest austenitizing temperature condition (Fig. 6(b)) is very

1900 Notch Tensile Strength ky = 14.1 MPa·mm0.5

1800

1700

Strength, MPa

10

Ultimate Tensile Strength ky = 8.7 MPa·mm0.5

1600

1500 1200

~

~ Yield Strength ky = 9.3 MPa·mm0.5

1100

1000 0.15

0.20

0.25

0.30

0.35

0.40

0.45

(PAGS)-1/2, m-1/2 Fig. 11. Hall-Petch plots of 0.2% offset yield strength and ultimate tensile strength obtained from the smooth-sided tensile tests and notch tensile strength obtained from the DENT tests as a function of the inverse square root of PAGS.

likely an influence of the coarse prior austenite grain structure. Moreover, the bend angle at maximum load measurements also generally decreased with increasing PAGS (Fig. 7(b)). Furthermore, the observation of quasi-cleavage fracture in the steel matrix of the specimens austenitized at the highest austenitizing temperature is evidence that the martensitic matrix with the largest PAGS had reduced toughness compared to the lower austenitizing temperature conditions (850 °C and 900 °C) (Figs. 9 and 10). These results are in agreement with previous reports that toughness of lath martensitic steels decreases as the PAGS and lath martensite packet size increase [5,32]. Wang et al. [32] reported that microstructural refinement in lath martensitic steels is beneficial for crack resistance because martensite packet boundaries can strongly hinder cleavage crack propagation. It has similarly been reported that reducing the PAGS has a beneficial influence on the ductile to brittle transition temperature in PHS as measured by stacked Charpy specimens [2]. However, it should be pointed out that the fracture mode in the martensitic PHS used in the present study was primarily ductile and involved microvoid nucleation and coalescence. While finer grain size is known to promote resistance to brittle fracture through increased crack deflection, the effect of grain size on ductile fracture is less clear in literature. In the microvoid nucleation and growth process, cracks do not deflect at grain boundaries in the same manner as brittle cracks. Therefore, open questions remain about the effect of the PAGS and substructure boundaries on ductile fracture of lath martensitic steels. There are some other factors that could have influenced the PHS toughness, albeit likely minor compared to the effect of PAGS such as presence of second phase particles. Specifically, Jo et al. [9] implicated a very small amount (1 vol%) of untransformed ferrite in the microstructure of a PHS for the deterioration of the impact toughness. However, in the present study, an unfavorable impact of untransformed ferrite (b1 vol%) on the toughness of Al-Si coated PHS was not observed for the smooth-sided tensile test or DENT test results (Figs. 5 and 6). 4.3. Microstructural variables affecting bendability In the DENT testing, PAGS was found to be the most important factor governing the mechanical and fracture behavior, as discussed in the previous section. While PAGS also played a role in the bending

L. Golem et al. / Materials and Design 172 (2019) 107707

performance of Al-Si coated PHS, the PAGS-dependence was less clear in the bending angle measurements compared to the notch displacement measurements (Figs. 6(b) and 7(b)). This could be due, in part, to the fact that some of the material parameters that relate to the bending performance, e.g. PAGS, coating microstructure, and microstructural homogeneity, change simultaneously with a change of austenitizing parameters. It has been shown that the bendability of ultra-high strength sheet steels is influenced by various material parameters such as the strength [33,34], grain size [34–36], surface condition (surface roughness, subsurface hardness, and subsurface microstructure) [37–39], microstructural homogeneity [33,39], and precipitates [40]. In general, bendability of a material decreases with increasing strength. However, Yamazaki et al. [33] argued that the bendability of ultra-high strength steels did not correlate with the tensile strength or ductility. In free bend testing, strain is greatest at the surfaces and therefore, it is reasonable to assume that the coating microstructure and thickness should affect the bending performance of an Al-Si coated PHS more significantly than the results of testing conducted in tension. According to the study of Windmann et al. [14], the substrate-coating interdiffusion layer for Al-Si coated PHS in the as-quenched state contains hard and brittle intermetallic phases, Al2Fe3Si3, Fe2Al5 and FeAl. In particular, Fe2Al5, the major constituent in the interdiffusion layer, was reported to have a low fracture toughness of approximately 1 MPa·m0.5 [41]. The presence of the brittle intermetallic phases in an Al-Si coating layer can cause a deterioration of the bending performance of PHS. The low toughness of the intermetallic phases is also evidenced here by the observation of transgranular or intergranular fracture features in the fracture zone of the coating (Fig. 10). It is noteworthy that the area of brittle intergranular or transgranular fracture zones in the bending specimens increased with increasing austenitizing temperature (Fig. 10), indicating an unfavorable influence of increased interdiffusion layer thickness. It is also important to recall the observation of quasicleavage fracture in the steel matrix of the bending specimen austenitized at the highest austenitizing temperature (Fig. 10(c)). Therefore, both the significant increase in the interdiffusion layer thickness and the reduced toughness of the steel associated with the coarse PAGS could have contributed to the substantial reduction in the bending angle at the highest austenitizing temperature condition (1025 °C30 min), shown in Fig. 7(b). Cracks formed at the surface may also act as a stress concentrator. A thicker coating layer, which is subject to brittle fracture, potentially results in the formation of longer cracks extending into the sheet during free bend testing. Furthermore, the higher austenitizing temperature seems to provide a thick adherent interdiffusion layer. In this case, the cracks that form in the brittle coating layer can propagate more easily into the substrate material instead of promoting coating spallation, which occurs for less adherent coatings. The sharp cracks that propagate through the coating and arrest at the steel substrate can cause a notch effect [42]; i.e. the cracks, which induce local stress concentrations, create a triaxial stress state on the substrate surface. As shown in the DENT experiments, the triaxial stress state can promote quasi-cleavage fracture, particularly in the condition austenitized at 1025 °C for 30 min (Fig. 10(c)). Therefore, the present study suggests that control of the thickness of the interdiffusion layer is also critical to the bending performance of Al-Si coated PHS. In the free bend test results (Fig. 7(a)), there was a noticeable effect of hold time on the bend angle for the specimens austenitized at 850 °C. The bending angle for the short hold times (3 min and 6.5 min) was low when compared to the longer hold time condition (850 °C-30 min). The reason for the observed hold time dependence of the bending angle is not well understood, given that the fraction of the untransformed ferrite was comparable between the specimens austenitized at 850 °C. Coarse precipitates or inclusions are known to be detrimental to bendability as they can promote nucleation of microvoids due to incompatibility between the matrix and particle

11

[40]. It is therefore possible that the presence of undissolved, coarse cementite particles in the specimens austenitized for the short hold times may have contributed to the decreased bendability. Further investigation involving quantitative analysis of the precipitates is needed to determine whether the precipitates have an impact on the bending performance of the Al-Si coated PHS. 5. Conclusions The key conclusions of the current study on mechanical behavior of Al-Si coated PHS are as follows: 1. Increasing the austenitizing temperature or hold time results in an increase in the PAGS for Al-Si coated PHS. When the steel was austenitized at 850 °C, slightly below the Ac3 temperature, the microstructure consisted of a lath martensitic matrix with small amounts of untransformed ferrite (b1 vol%). The microstructure of the steel austenitized above the Ac3 temperature was fully martensitic. The mechanical behavior was not substantially influenced by the small amounts of untransformed ferrite. 2. The stress and strain state experienced during testing affects the sensitivity of the mechanical and fracture behavior of Al-Si coated PHS to austenitizing conditions. That is, the tensile elongation was relatively independent of the austenitizing conditions, while there was a noticeable influence of the austenitizing conditions on the notch displacement and the bending angle results. 3. Refinement of the PAGS was, in general, beneficial for the mechanical performance of Al-Si coated PHS. The fracture of the martensitic matrix was usually ductile. However, quasi-cleavage fracture was observed in the martensitic matrix with the largest PAGS, which is due to a loss of toughness, possibly associated with increased substructure size. Strength, notch displacement, notch strength ratio, percent flat fracture, bending maximum load, and bend angle at maximum load all generally decreased with increasing austenitizing temperature and hold time, i.e. increasing PAGS. In particular, yield strength, smooth specimen tensile strength, and notched specimen strength increased in inverse proportion to the square root of the PAGS. 4. The thickness of the substrate-coating interdiffusion layer increased with increasing austenitizing temperature or hold time. The interdiffusion layer of the Al-Si coated PHS in the as-quenched state contains FeAl intermetallic crystals that are more brittle as compared to the martensitic matrix, evidenced by the observation of the transgranular or intergranular fracture features in the fracture zone of the coating. Thicker interdiffusion layers developed at higher austenitizing temperature, negatively influencing bendability, i.e. bend angle at maximum load, of the material. CRediT authorship contribution statement Lindsay Golem: Conceptualization, Formal analysis, Investigation, Methodology, Writing - original draft, Writing - review & editing. Lawrence Cho: Conceptualization, Formal analysis, Investigation, Methodology, Writing - original draft, Writing - review & editing. John G. Speer: Conceptualization, Formal analysis, Funding acquisition, Resources, Writing - review & editing. Kip O. Findley: Conceptualization, Formal analysis, Methodology, Funding acquisition, Resources, Supervision, Writing original draft, Writing - review & editing. Acknowledgments The authors gratefully acknowledge the support and technical insight of the sponsors of the Advanced Steel Processing and Products Research Center at the Colorado School of Mines. The authors greatly appreciate the support of Gestamp HardTech for providing and processing all the material used for this investigation.

12

L. Golem et al. / Materials and Design 172 (2019) 107707

Data availability statement The raw and processed data to reproduce these findings are available upon request. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.matdes.2019.107707. References [1] K.O. Findley, L. Golem, L. Cho, K.D. Clarke, Evaluating Crash Relevant Properties of Advanced High Strength SteelsProceedings of Asia Steel 2018 Conference, India, 2018. [2] J. Wang, C.M. Enloe, J.P. Singh, C. Horvath, Effect of prior austenite grain size on impact toughness of press hardened steel, SAE Int. J. Mater. Manuf. 9 (2016) 488–493. [3] P. Larour, J. Naito, A. Pichler, T. Kurz, T. Murakami, Side Impact Crash Behavior of Press-Hardened Steels-Correlation With Mechanical PropertiesProceedings of 5th International Conference on Hot Sheet Metal Forming of High-Performance Steel (CHS2 2015), Toronto, Canada 2015, pp. 281–289. [4] C.M. Enloe, J. Wang, J.P. Singh, C. Horvath, N. Ramisetti, S. Sriram, Process Influences on Press-Hardened Steel Microstructure and Impact PerformanceProceedings of the Iron & Steel Technology Conference (AISTech 2016), Pittsburgh, PA 2016, pp. 16–19. [5] H. Hoseiny, F.G. Caballero, D. San Martin, C. Capdevila, The influence of austenitization temperature on the mechanical properties of a prehardened mould steel, Mater. Sci. Forum (2012) 2140–2145 Trans Tech Publ. [6] A. Andreiev, O. Grydin, M. Schaper, Evolution of microstructure and properties of steel 22MnB5 due to short austenitization with subsequent quenching, Steel Res. Int. 87 (12) (2016) 1733–1741. [7] H. Cai, P. Du, H. Yi, D. Wu, Effects of austenitizing temperature on microstructure and properties of hot-formed steel, Adv. Mater. Res. 1063 (2014). [8] W.S. Choi, Characterization of the Bendability of Press Hardened 22MnB5 Steel, (M.S. Thesis) Pohang University of Science and Technology, South Korea, 2013. [9] M.C. Jo, J. Park, S.S. Sohn, S. Kim, J. Oh, S. Lee, Effects of untransformed ferrite on charpy impact toughness in 1.8-GPa-grade hot-press-forming steel sheets, Mater. Sci. Eng. A 707 (2017) 65–72. [10] S. Otani, M. Kozuka, T. Murakami, J. Naito, A. Pichler, T. Kurz, Metallurgical Controlling Factors for the Ductility of Hot Stamped PartsProceedings of 5th International Conference on Hot Sheet Metal Forming of High-Performance Steel (CHS2 2015), Toronto, Canada 2015, pp. 411–416. [11] J. Kondratiuk, P. Kuhn, Tribological investigation on friction and wear behaviour of coatings for hot sheet metal forming, Wear 270 (11−12) (2011) 839–849. [12] Z.-x. Gui, W.-k. Liang, Y. Liu, Y.-s. Zhang, Thermo-mechanical behavior of the Al–Si alloy coated hot stamping boron steel, Mater. Des. 60 (2014) 26–33. [13] L. Dosdat, J. Petitjean, T. Vietoris, O. Clauzeau, Corrosion resistance of different metallic coatings on press-hardened steels for automotive, Steel Res. Int. 82 (6) (2011) 726–733. [14] M. Windmann, A. Röttger, W. Theisen, Phase formation at the interface between a boron alloyed steel substrate and an Al-rich coating, Surf. Coat. Technol. 226 (2013) 130–139. [15] Z.-X. Gui, K. Wang, Y.-S. Zhang, B. Zhu, Cracking and interfacial debonding of the Al– Si coating in hot stamping of pre-coated boron steel, Appl. Surf. Sci. 316 (2014) 595–603. [16] ASTM E112-96, Standard Test Methods for Determining Average Grain Size, ASTM International, 2004. [17] L. Golem, Influence of Austenitizing Parameters on Mechanical Behavior of Press Hardened Steels, (M.S. Thesis) Colorado School of Mines, Golden, CO, 2017.

[18] VDA 238-100, Test Specification Draft: Plate Bending Test for Metallic Materials, 12/ 2010. [19] B. Hutchinson, J. Hagström, O. Karlsson, D. Lindell, M. Tornberg, F. Lindberg, M. Thuvander, Microstructures and hardness of as-quenched martensites (0.1–0.5% C), Acta Mater. 59 (14) (2011) 5845–5858. [20] K.-R. Jo, E.-J. Seo, D.H. Sulistiyo, J.-K. Kim, S.-W. Kim, B.C. De Cooman, On the plasticity mechanisms of lath martensitic steel, Mater. Sci. Eng. A 704 (2017) 252–261. [21] C. Georges, T. Sturel, P. Drillet, J.-M. Mataigne, Absorption/desorption of diffusible hydrogen in aluminized boron steel, ISIJ Int. 53 (2013) 1295–1304. [22] P. Larour, B. Hackl, F. Leomann, Sensitivity Analysis on the Calculated Bending Angle in the Instrumented Bending TestProc. Int. Conf. IDDRG 2013, pp. 02–05. [23] M.A. Crimp, K. Vedula, Effect of boron on the tensile properties of B2 FeAl, Mater. Sci. Eng. 78 (2) (1986) 193–200. [24] C.C. Roth, D. Mohr, Ductile fracture experiments with locally proportional loading histories, Int. J. Plast. 79 (2016) 328–354. [25] G. Pape, Flow Stress and Ductile Failure at Varying Strain Rates, (Ph.D. Thesis) Delft University of Technology, Delft, The Netherlands, 2003. [26] R. Qu, P. Zhang, Z. Zhang, Notch effect of materials: strengthening or weakening? J. Mater. Sci. Technol. 30 (6) (2014) 599–608. [27] A. Asserin-Lebert, J. Besson, A.-F. Gourgues, Fracture of 6056 aluminum sheet materials: effect of specimen thickness and hardening behavior on strain localization and toughness, Mater. Sci. Eng. A 395 (1–2) (2005) 186–194. [28] M. Faccoli, G. Cornacchia, M. Gelfi, A. Panvini, R. Roberti, Notch ductility of steels for automotive components, Eng. Fract. Mech. 127 (2014) 181–193. [29] F. Bron, J. Besson, A. Pineau, Ductile rupture in thin sheets of two grades of 2024 aluminum alloy, Mater. Sci. Eng. A 380 (1–2) (2004) 356–364. [30] S. Morito, H. Yoshida, T. Maki, X. Huang, Effect of block size on the strength of lath martensite in low carbon steels, Mater. Sci. Eng. A 438 (2006) 237–240. [31] S.C. Kennett, Strengthening and Toughening Mechanisms in Low-C Microalloyed Martensitic Steel as Influenced by Austenite Conditioning, (Ph.D. Thesis) Colorado School of Mines, Golden, CO, 2014. [32] C. Wang, M. Wang, J. Shi, W. Hui, H. Dong, Effect of microstructural refinement on the toughness of low carbon martensitic steel, Scr. Mater. 58 (6) (2008) 492–495. [33] K. Yamazaki, M. Oka, H. Yasuda, Y. Mizuyama, H. Tsuchiya, Recent Advances in Ultrahigh-Strength Sheet Steels for Automotive Structural Use, Nippon Steel Technical Report No. 64, 1995 37–44. [34] B. Jian, L. Wang, H. Mohrbacher, H.Z. Lu, W.J. Wang, Development of niobium alloyed press hardening steel with improved properties for crash performance, Adv. Mater. Res. (2015) 7–20 Trans Tech Publ. [35] J. Tu, K.-C. Yang, L.-J. Chiang, W. Cheng, The Effect of Niobium and Molybdenum Coaddition on Bending Property of Hot Stamping Steels, China Steel Technical Report No. 29, 2016 1–7. [36] H. Mohrbacher, Reverse metallurgical engineering towards sustainable manufacturing of vehicles using Nb and Mo alloyed high performance steels, Adv. Manuf. 1 (2013) 28–41. [37] W.S. Choi, B.C. De Cooman, Characterization of the bendability of press-hardened 22MnB5 steel, Steel Res. Int. 85 (2014) 824–835. [38] A.J. Kaijalainen, P. Suikkanen, L.P. Karjalainen, J.J. Jonas, Effect of austenite pancaking on the microstructure, texture, and bendability of an ultrahigh-strength strip steel, Metall. Mater. Trans. A 45 (2014) 1273–1283. [39] A.J. Kaijalainen, P.P. Suikkanen, L.P. Karjalainen, D.A. Porter, Influence of subsurface microstructure on the bendability of ultrahigh-strength strip steel, Mater. Sci. Eng. A 654 (2016) 151–160. [40] J. Steninger, A. Melander, The relation between bendability, tensile properties and particle structure of low-carbon steel, Scand. J. Metall. 11 (2) (1982) 55–71. [41] U. Köster, W. Liu, H. Liebertz, M. Michel, Mechanical properties of quasicrystalline and crystalline phases in Al-Cu-Fe alloys, J. Non-Cryst. Solids 153 (1993) 446–452. [42] J. Schaufler, K. Durst, O. Massler, M. Göken, In-situ investigation on the deformation and damage behaviour of diamond-like carbon coated thin films under uniaxial loading, Thin Solid Films 517 (5) (2009) 1681–1685.