Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels

Engineering Fracture Mechanics xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Engineering Fracture Mechanics journal homepage: www.elsevier.com/locate/engfracmech

Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels Fayaz Foroozmehr a,⇑, Yves Verreman a, Jianqiang Chen a, Denis Thibault b, Philippe Bocher c a

École Polytechnique de Montréal, C.P.6079, succ. Centre-ville, Montréal, Québec H3C 3A7, Canada Institut de Recherche d’Hydro-Québec, 1800, boul. Lionel-Boulet, Varennes, Québec J3X 1S1, Canada c École de Technologie Supérieure (ÉTS), Montréal, Québec H3C 1K3, Canada b

a r t i c l e

i n f o

Article history: Received 9 April 2016 Received in revised form 3 February 2017 Accepted 5 February 2017 Available online xxxx Keywords: Fracture toughness Microstructure Martensitic stainless steels Inclusions TRIP effect

a b s t r a c t Microstructures and inclusion characteristics of cast and wrought 13% Cr-4% Ni martensitic stainless steels were related to tensile, fracture toughness JIc, and Charpy V-notch absorbed energy. A lower volume fraction of inclusions, a larger number of inclusions smaller than 10 mm, and a more resistant inclusion type against rupture and against micro-void formation were the key parameters explaining why the wrought steel has better mechanical properties than the cast steel. It has been shown that TRIP effect occurs only in a small region of 6 mm length from the crack plane in JIc-test. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Since the 1960s, different grades of soft martensitic stainless steels such as 13% Cr-1.5% Ni, 13% Cr-4% Ni, 16% Cr-6% Ni and 17% Cr-4% Ni have been produced [1]. Among these grades, 13% Cr-4% Ni steel is used to fabricate hydraulic turbine runners because of its good corrosion resistance, good resistance to cavitation erosion, excellent weldability, and good castability [2–4]. This steel is also used in compressor cones, pump bowls, and petrochemical industries [5]. Delta ferrite is known to have deleterious effects on the mechanical properties of chromium martensitic stainless steels, especially on fracture toughness and absorbed impact energy [5–7]. This can be caused by carbide precipitation at delta ferrite grain boundaries during heat treatment; the precipitates acting as crack initiation and propagation sites [8,9]. As a consequence, nickel was added to the chromium martensitic stainless steels to remove the delta ferrite and improve toughness [1,10]. However, often, some supercooled delta ferrite can be found along prior austenite grains; especially in the case of cast martensitic stainless steels. This is due to the segregation of ferrite-promoting elements during solidification which stabilizes the delta ferrite at room temperature. In the as-quenched state, the microstructure of 13% Cr-4% Ni martensitic stainless steels is a combination of lath martensite and small amounts of delta ferrite. During subsequent tempering at around 600 °C, the martensite is partially transformed into stable austenite, forming the so-called ‘‘reformed austenite” phase. This reformed austenite is thermally stable even if the steel is quenched in liquid nitrogen (
196 °C) [5]. On the other hand, tempering at higher temperatures would lead to the formation of thermally unstable austenite which turns into fresh martensite during cooling [1,8].

⇑ Corresponding author. E-mail address: [email protected] (F. Foroozmehr). http://dx.doi.org/10.1016/j.engfracmech.2017.02.002 0013-7944/Ó 2017 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

2

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

Nomenclature a0 inclusion size E Young’s modulus elongation at fracture ef eu uniform elongation (before necking) f volume fraction of inclusions H(d) harmonic mean value of inclusion diameters J-integral value at the initiation of stable crack growth JIc J-R curve J-Resistance curve KIc plane strain stress intensity factor T tearing modulus a stress concentration factor around an inclusion cre reformed austenite cs surface energy Da crack extension k inclusion spacing 0.2% ry 0.2% offset yield strength ry yield strength rUTS tensile strength rvoid required stress for void nucleation Acronyms ACI Alloy Casting Institute CT Compact Tension specimen CVN Charpy V-Notch EDX Energy Dispersive X-ray FEM Finite Element Modeling LEFM Linear Elastic Fracture Mechanics MVC Micro-Void Coalescence RD Rolling Direction SEM Scanning Electron Microscope TD Transverse Direction UNS Unified Numbering System XRD X-Ray Diffraction

The reformed austenite is mechanically unstable, and transforms into martensite under applied deformation [4,5]. This transformation-induced plasticity (TRIP) is an important mechanism which contributes to the high mechanical properties of 13% Cr-4% Ni martensitic stainless steels. It retards necking during uniaxial loading by increasing the strain hardening of the alloy leading to a significant increase of the material strength. Bilmes et al. [2,5] studies showed that a microstructure without delta ferrite, consisting of tempered martensite and uniformely dispersed reformed austenite results in a good combination of strength and ductility. During crack propagation, the strain energy of the crack tip leads to the transformation of this metastable austenite into martensite and part of the energy which is needed to create new surfaces is absorbed by this transformation increasing the material toughness. Furthermore, compressive stresses produced by the volumetric expansion (about 4%) made by the formation of the martensite phase can close and deflect the crack tip [11]. It has been shown by Thibault et al. [4] that TRIP effect also takes place during fatigue crack propagation in cast and wrought 13% Cr-4% Ni martensitic stainless steels. Density, spacing, and size of inclusions are the principle parameters that affect micro-void formation in dimpled rupture. Lower density and higher spacing of inclusions provide less nucleation sites for micro-void formation, improving fracture toughness [12–14]. Inclusion size plays also an important role in dimpled rupture. Investigations made by Gurland and Plateau [15] showed that as inclusion size increases, the stress needed for void nucleation decreases by the following equation:

rv oid ¼

 0:5 1 Ecs a a0

ð1Þ

where rvoid is required stress for void nucleation, a is stress concentration factor around the inclusion, cs is the surface energy needed to create new surfaces and a0 is the inclusion size. Micro-voids initiate by matrix-inclusion decohesion and/or by inclusion rupture. In most cases, large inclusions fracture at low strains [16]. Further, large inclusions generally Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

3

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

fracture at low strains [16], facilitating the formation of micro-voids. Eriksson [17] proposed that fracture toughness is affected more by the inclusion spacing in plane strain state, whereas it is the inclusion size which controls the fracture process in plane stress state conditions due to the resulting large plastic zone. Based on the authors’ knowledge, despite of important papers and basic books such as [12,15,16,18–20], no comprehensive investigation has been reported so far on the effects of the microstructural features including distribution, size, and types of inclusions on the mechanical properties of cast and wrought 13% Cr-4% Ni martensitic stainless steels, especially for the wrought one. In this study three mechanical tests (tensile, JIc, and Charpy impact tests) were performed on one cast and one wrought martensitic 13% Cr-4% Ni stainless steel, both steels in tempered condition. In addition, TRIP effect of austenite during JIc testing was examined using XRD, and the TRIP effect zone was determined in a direction perpendicular to the crack plane for the wrought steel. 2. Materials and experimental procedures 2.1. Material CA6NM is the cast version and S41500 is the wrought version of martensitic 13% Cr-4% Ni stainless steels. CA6NM nomenclature is defined by ACI (Alloy Casting Institute) and that of S41500 refers to the UNS (Unified Numbering System). In this study, the S41500 steel nomenclature is simplified as 415 steel. For the microstructural and mechanical characterization of CA6NM steel, all specimens were taken randomly. In the case of 415 steel, the specimen orientation is given in the Sections 2.2 and 2.3. In addition, rolling and transverse directions were shown on the micrographs of the microstructure and fracture surfaces of 415 steel. The chemical compositions of the two materials studied in this paper are shown in Table 1. Their composition requirements according to ASTM A743 [21] and ASTM A240 [22] are also added. All of the chemical composition contents verify these standard requirements. The CA6NM and 415 steels have been austenitized at 1050 °C and 1000 °C then tempered at 610 °C and 600 °C, respectively, followed by air quenching. 2.2. Microstructure In order to observe the microstructures by optical microscopy, samples were ground using sand papers and diamond suspensions. Metallographic sections of as-received materials were examined before etching to determine the surface density, volume fraction, size, and spacing of inclusions. Measurements were made at a magnification of 50 on polished surfaces and a total surface of 14 mm2 was examined for each material. This low magnification level was used to have a more precise idea about the distribution of large inclusions, since they highly affect the inclusion volume fraction and the mechanical properties. At least 150 inclusions were studied for each material. The measurement process was performed using an auto
1=3

matic image analysis software. Inclusion average spacing k was calculated using the relationship k ¼ 0:89R0 f where R0 ¼ ðp=4ÞHðdÞ, H(d) being the harmonic mean value of inclusion diameters and f is the volume fraction of inclusions [13]. A modified Fry’s etching solution (150 mL distilled water, 1 g CuCl2, 50 mL HCl, and 50 mL HNO3) was used to reveal the microstructures and measure prior austenite grain size [4]. Samples were slightly etched to reveal the prior austenite grain boundaries. The grain sizes were measured in T-S, S-L, and L-T planes in the case of 415 steel since there can be some anisotropy induced by the rolling process. Equivalent circle diameters were calculated from measured grain areas and an arithmetic mean of 100 grain diameters was defined as the prior austenite grain size for each material/orientation. To this purpose, 15 optical micrographs at 100 and 45 ones at 200 were examined for CA6NM steel and 415 steel, respectively. 2.3. Mechanical testing Tensile tests were carried out at room temperature according to ASTM E8/E8M [23] using cylindrical specimens of 5 mm diameter. In order to determine the possible anisotropy of 415 steel due to the rolling process, the tests were performed in the rolling and transverse directions (L, T). Three tests were performed for each material/orientation. The fracture toughness JIc of the studied materials was measured using side-grooved compact tension (CT) specimens by the single-specimen test method described in ASTM E1820 [24]. The thickness of the specimens was 25.4 mm and the width

Table 1 Chemical compositions and ASTM requirements for CA6NM and 415 steels. Designation

C

Mn

P

S

Si

Cr

Ni

Mo

CA6NM steel ASTM A743 [19]

0.016 0.06 (max) 0.028 0.050 (max)

0.50 1.00 (max) 0.73 0.50–1

0.019 0.04 (max) 0.022 0.030 (max)

0.007 0.03 (max) 0.002 0.030 (max)

0.26 1.00 (max) 0.34 0.60 (max)

12.79 11.5–14.0

3.49 3.5–4.5

0.55 0.40–1.0

13.021 11.5–14

3.91 3.5–5.5

0.56 0.5–1

415 steel ASTM A240 [20]

Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

4

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

was 50.8 mm. The depth of the side-grooves on each side of the specimens was 10% of the whole thickness. Crack front straightness and crack lengths have been verified on the fracture surfaces. More details on the fracture toughness testing of the studied materials are provided in [25]. Three tests were performed for each material at room temperature. In the case of 415 steel, specimens were machined in L-T orientation and one more specimen was machined in the T-L orientation to have an idea if there is some anisotropy in fracture toughness. Notched bar specimens were used for CVN tests carried out according to ASTM E23 [26]. To determine the ductile to brittle transition curves three tests were performed at eight temperatures for both materials, from
190 °C to + 80 °C. In the case of 415 steel; the tests were performed using specimens with L-T orientation. 2.4. Fractography and X-ray diffraction analysis A scanning electron microscope (SEM) equipped with an energy dispersive X-ray (EDX) spectrometer was used to investigate fracture surfaces and chemical nature of inclusions. SEM micrographs which were representative of all regions observed on the fracture surface for each material/test were chosen to discuss in this paper. More than 270 SEM micrographs at different magnifications and 147 EDX spectrums were examined. The inclusion size was also measured on the fracture surfaces of tensile, fracture toughness JIc, and CVN specimens tested at room temperature. Since automatic measurements were not possible in these cases; magnifications of more than 500 were used to increase the precision of these manual measurements. More than 100 inclusions were measured for each steel fracture surfaces. Only intact inclusions which were considered as representative of all ones existed on the fracture surfaces were examined (fractured inclusions and those overlapping dimple walls were ignored). In order to examine the contribution of the TRIP effect on the mechanical properties, austenite contents were measured on metallographic sections of as-received materials as well as perpendicular to the JIc-test crack plane by means of quantitative X-ray diffraction (XRD). Diffraction peak intensities were evaluated using the Rietveld method [27]. A diffractometer equipped with a copper tube and a nickel filter was used. Diffraction patterns were obtained in the range 40° < 2h < 140° with a step size of 0.05°. Other experimental details regarding the quantitative austenite measurement procedures were reported in [4]. All of the samples used for austenite measurements were cut by an automatic cut-off machine. A feed speed of 0.05 mm/s and 0.5 mm thick saw blades were used to prevent transformation of austenite during the cutting process. Cut samples were ground by 1200 grit sand papers and etched in a 30% HCl + 30% HNO3 + 40% H2O solution for 10 min. These samples were etched again in HCl for 30 s. The surfaces of metallographic sections were 63 mm2 and 84 mm2 for CA6NM steel and 415 steel, respectively. Three samples were prepared from each material to measure the initial austenite content. To quantify the amount of TRIP in the fractured JIc-test specimens, two 20  10  2 mm samples were cut (one from the cast steel and one from the wrought steel) as represented in Fig. 1(a). A surface of 7 mm2 was large enough to properly diffract each austenite measurement region shown in Fig. 1(b). 3. Results 3.1. Microstructure The typical microstructure of the lath martensite can be observed in Fig. 2 for both steels. Packets and blocks of martensite lathes can be seen in both microstructures. As shown in this figure, CA6NM steel has a coarse microstructure as compared to 415 steel. This is confirmed by the prior austenite grain size measurements provided in Table 2. The prior austenite grain size of CA6NM steel is approximately 85% larger than that of 415 steel. The prior austenite grain size of 415 steel is approximately equal for the three T-S, S-L and L-T cutting planes. It seems not be affected by the rolling process and the grains can be considered as equiaxed. Both microstructures shown in Fig. 2(a) and (b) consist of tempered martensite and finely-dispersed islands of reformed austenite. The austenite phase in the martensite matrix is so fine that it cannot be observed by optical microscopy; higher magnifications are needed. Investigations made on the same materials but from different batches [4,28] showed that the austenite phase is in the form of lamellae located within the laths of martensitic matrix. The initial austenite contents of CA6NM and 415 steels measured on metallographic sections are reported in Table 3. Both values can be considered as equal, however, the standard deviation is relatively high in the case of CA6NM steel. There is a heterogeneous distribution of reformed austenite that originates from heterogeneous concentrations of austenite-promoting elements in the cast microstructure. Delta ferrite was observed only in the CA6NM samples. This is illustrated in Fig. 3(a) and (b) using 50 and 100 magnifications, respectively. Parallel elongated traces of delta ferrite are displayed in (a) while stringers of delta ferrite are observed along the prior austenite grains in (b). The surface density, volume fraction, mean diameter, and inclusion average spacing of CA6NM and 415 steels are shown in Table 4. The surface density of inclusions is about 35% higher for CA6NM steel, whereas its volume fraction is more than twice as much as that of 415 steel. The calculated inclusion spacings in both steels are approximately equal. Size distributions of inclusions measured on the metallographic sections of the studied materials are shown in Fig. 4. The relative frequency of inclusions having diameters of more than 10 mm in CA6NM steel is more than twice as much as those in Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

5

Fig. 1. (a) Position of the cut sample (dotted rectangular prism) in a JIc fractured specimen and (b) austenite (c) measurement regions on the cut sample, as shown by the arrows.

415 steel (31% vs. 14%). These large inclusions are three times more numerous in CA6NM steel if the absolute frequencies are compared.

3.2. Tensile behavior 3.2.1. Tensile properties The tensile properties of the two steels and ASTM requirements at room temperature are shown in Table 5. All of the results substantiate the standard requirements. The ductility at fracture is slightly higher for 415 steel. Properties in the L- and T- orientation are similar and no strong anisotropy is observed.

3.2.2. Fractography of tensile tests Typical fracture surfaces of CA6NM and 415 steels tested at room temperature are shown in Fig. 5(a) and (b), respectively. Note that the observed surface is T-S in the case of 415 steel. Both fracture surfaces represent dimpled rupture, but their appearances are different. In the case of CA6NM steel, the fracture surface consists only of dimples, while the void sheets are also observed in 415 steel. In Fig. 5(b), a secondary crack was formed due to the coalescence of some micro-voids. This crack propagated and connected the other micro-voids by void sheets formation. The inclusion size measured on the fracture surfaces is provided in Table 6. The mean diameter is 25% larger than that measured on metallographic sections (Table 4) in the case of CA6NM steel. Contrarily, inclusions of half as large as those measured on metallographic sections are found on the 415 steel fracture surfaces. The size distributions of inclusions measured on the tensile test fracture surfaces of CA6NM and 415 steels are shown in Fig. 6. Fifty-five percent of inclusions are of more than 10 mm in diameter in the case of CA6NM steel, whereas the relative frequency of these large inclusions is only 3% in the case of 415 steel. On the other hand, it is interesting to mention that all inclusions detected on 415 steel fracture surfaces were intact, whereas some were fractured on CA6NM steel fracture surfaces. It shows that there is only one operating mechanism for Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

6

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

Fig. 2. Optical micrographs of (a) CA6NM steel and (b) 415 steel.

dimple nucleation on 415 steel i.e., decohesion of matrix-inclusion, but another source of micro-void nucleation on CA6NM steel is the fracture of inclusions. 3.3. Fracture toughness 3.3.1. JIc and CVN properties Fig. 7 shows the J resistance curves (J-R curves) at room temperature. In the case of 415 steel, the one of L-T direction is shown. The following equation was used to evaluate the construction line:

J ¼ 2rY Da

ð2Þ

where

rY ¼

ry þ rUTS

ð3Þ

2

Table 2 Prior austenite grain size of the studied materials. Material

Observed surface

Mean grain size (mm)

CA6NM steel 415 steel

– (T-S) (S-L) (L-T)

163.28 ± 67.20 90.58 ± 34.68 88.94 ± 40.10 85.89 ± 38.14

Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

7

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx Table 3 Initial austenite materials.

contents

in

the

studied

Designation

Reformed austenite (%)

CA6NM steel 415 steel

19.1 ± 6.4 19.5 ± 2.2

Fig. 3. Optical micrographs showing delta ferrite traces in CA6NM steel: (a) Parallel elongated traces at 50 and (b) stringers along the prior austenite grain boundaries at 100.

and ry , and rUTS are yield, and tensile strengths, respectively, measured from the uniaxial tensile test. J ¼ C Dap formula was used to evaluate the J-R curves, and data points intercepted between 0.15 mm and 1.5 mm exclusion lines were collected for this purpose. The fracture toughness candidate values, JQ, were obtained from the intersection of 0.2 mm offset lines and the J-R curves. After verifying the straightness of both fatigue pre-crack and JIc crack extension fronts, as well as ASTM E1820 validity requirements, the JQ values were qualified as JIc since all the standard requirements were satisfied.

Table 4 Inclusion characteristics of the studied materials. Designation

Observed surface

Surface density (No./cm2)

Volume fraction (%)

Size (mean diameter, mm)

k (mm)

CA6NM steel 415 steel

– (T-S)

1630 ± 368 1210 ± 242

0.22 ± 0.12 0.09 ± 0.02

9.98 ± 8.62 7.18 ± 6.42

37.07 ± 5.73 36.05 ± 3.68

Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

8

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

Fig. 4. Size distributions of inclusions measured on metallographic sections.

Plane strain fracture toughness properties, CVN upper shelf impact energies, and tensile strengths at room temperature of the studied materials are given in Table 7. Fracture toughness values of 415 steel correspond to the tests performed in L-T orientation. The slopes of the J-R curves were calculated using the parts of the curves located between the exclusion lines. Additionally, since JIc is a measure of fracture toughness at initiation of stable crack growth in ductile behaviors, the entire resistance of a material against stable crack growth cannot be explained by this initiation fracture toughness. Accordingly, a more appropriate criterion explaining the resistance of a ductile material against stable crack growth, entitled ‘‘tearing modulus”, was defined for this purpose [16]:

T¼

E dJ

ð4Þ

r2y da

The tearing modulus is a dimensionless criterion which is related directly to the slope of the J-R curve. As shown in Fig. 7, both steels have similar J-R curves, but that of 415 steel is steeper. As provided in Table 7, the slope of J-R curve is 44% higher for 415 steel. Inversely, its tearing modulus is about nine percent lower than that of CA6NM steel. Both steels have high fracture toughness and high upper shelf CVN absorbed energies. However, 415 steel which has the highest ultimate tensile strength, has the highest CVN absorbed energy and fracture toughness JIc. Another JIc test carried out in T-L orientation, gave a fracture toughness JIc value of 437 kJ/m2. Therefore, similar to the tensile properties, no significant anisotropy is observed between L-T and T-L in terms of fracture toughness. CVN ductile-brittle transition curves of the two steels are shown in Fig. 8. Both steels show approximately similar transition and lower shelf behaviors but large differences are observed in the upper shelf region. The upper shelf CVN absorbed energy is 49% higher for 415 steel. In a similar way, the fracture toughness JIc of 415 steel is 39% higher than that of CA6NM steel, as provided in Table 7. 3.3.2. Spread of the TRIP effect The evolution of austenite content with the distance from JIc fracture surface of 415 steel is shown in Fig. 9. It confirms the presence of the TRIP effect as it was already reported in [1,2,4,5,11] for 13% Cr-4% Ni martensitic stainless steels. The solid line shows the initial austenite content measured on metallographic sections of the as-received material and the dashed lines show its maximum and minimum values. Measurements have been taken every two millimeters until a constant value (the initial austenite content of the material) is reached. This determines the region where the austenite had transformed into martensite by TRIP effect. Three measurements have been done at a given distance.

Table 5 Tensile properties and ASTM requirements for CA6NM and 415 steels at room temperature. Designation

0.2% ry (MPa)

rUTS (MPa)

eu (%)

ef (%)

Reduction of area (%)

CA6NM steel ASTM A743 (min) [19] 415 steel (L) 415 steel (T) ASTM A240 (min) [20]

575 ± 13 550 725 ± 0 705 ± 9 620

769 ± 7 755 843 ± 4 835 ± 7 795

9.5 ± 0.4 – 9.3 ± 0.4 8.6 ± 0.6 –

19.4 ± 1.7 15 21.7 ± 1.0 19.8 ± 0.3 15

63.0 ± 3.8 – 70.4 ± 1.8 71.0 ± 1.4 –

Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

9

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

Fig. 5. Tensile test fracture surfaces of (a) CA6NM steel and (b) 415 steel; some intact/broken inclusions and micro-voids are shown by the arrows.

In the case of CA6NM steel, due to a large variation in the measurements, no clear plateau was found and the estimation of the TRIP effect region was not possible. On the other hand, the homogeneous distribution of the reformed austenite in the microstructure of 415 steel provides the opportunity to estimate the extent of the region where the austenite had transformed into martensite by TRIP. As shown in Fig. 9, the values measured from 6 mm to 14 mm correspond to the initial austenite content of 415 steel and, it can be concluded that TRIP effect occurs down to a depth of 6 mm. The plastic zone ahead of the crack tip calculated by Linear Elastic Fracture Mechanics (LEFM) is about 20 mm in length and 100 mm in height (plastic ears in plane strain) [16]. However, LEFM cannot be applied in the present case since small scale yielding conditions are violated (the specimen height is 61 mm). Elastic-plastic FEM analysis could be performed in order to determine whether the TRIP effect occurs in the whole plastic zone or in a part of it.

3.3.3. Fractography of fracture toughness JIc and CVN tests Typical regions from the JIc fracture surfaces of CA6NM and 415 steels, tested at room temperature, are shown in Fig. 10 (a) and (b), respectively. Both fracture surfaces are typical of dimpled rupture. The dimples are significantly larger in CA6NM steel. Broken inclusions as well as intact ones are observed on the CA6NM steel fracture surface, whereas all inclusions observed on the 415 steel fracture surface have been found to be intact.

Table 6 Inclusion size measured on fracture surfaces of tensile test samples. Material

Observed surface

Number of measured inclusions

Size (mean diameter, mm)

CA6NM steel 415 steel

– (T-S)

49 30

12.47 ± 8.30 3.60 ± 2.24

Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

10

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

Fig. 6. Size distributions of inclusion measured on tensile tests fracture surfaces.

Fig. 7. J-R curves of (a) CA6NM steel, and (b) 415 steel.

Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

11

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx Table 7 Plane strain fracture toughness JIc and CVN impact energies at room temperature of the studied materials.

a b

Material

JIc (kJ/m2)

dJ/da (MPa)

T

CVN (J)

rUTS (MPa)

CA6NM steel 415 steel

299 ± 13 416 ± 71a

153 ± 2 221 ± 4a

92 ± 1 84 ± 1a

142 ± 3 211 ± 6a

769 ± 7 843 ± 4b

Measured in L-T orientation. Measured in L orientation.

Fig. 8. CVN ductile-brittle transition curves of CA6NM and 415steels.

Regions from the CVN fracture surfaces of the tested materials at 23 °C are shown in Fig. 11(a) and (b). The fracture surfaces are very similar to the JIc ones in terms of inclusion size and dimples nucleation. As provided in Table 8, the inclusion size is also similar between JIc and CVN fracture surfaces for a given material; however, the inclusion size is at least 3 times larger in CA6NM steel. Size distributions of inclusion sizes measured on the JIc and CVN tests fracture surfaces performed at room temperature are shown in Fig. 12(a) and (b), respectively. In the similar manner as the tensile tests, inclusions of more than 10 mm in diameter were much more numerous on JIc and CVN fracture surfaces of CA6NM steel (63% and 79%, respectively) than those on JIc and CVN fracture surfaces of 415 steel (2% and 0%, respectively). A comparison between these inclusion sizes measured on JIc and CVN fracture surfaces, and those measured on metallographic sections (Table 4) shows that the dimples are mostly nucleated from larger inclusions on CA6NM steel, whereas smaller inclusions act as nucleation sites for 415 steel dimples suggesting that the density of large inclusions may control the fracture process.

Fig. 9. Evolution of c content from JIc fracture surface of 415 steel.

Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

12

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

Fig. 10. JIc fracture surfaces of (a) CA6NM steel and (b) 415 steel; some intact/broken inclusions are shown by the arrows.

In Fig. 11, micrographs (c) to (f) are also added to study the mechanisms involved in the tests carried out at
80 °C and
190 °C. Semi-cleavage (formation of dimples + cleavage rupture) and pure cleavage fractures are found in both steels, respectively. For semi-cleavage fracture mode (micrographs (c) and (d)), larger inclusions are observed again on the CA6NM steel fracture surface.

3.4. Inclusion types All inclusions observed on 415 steel fracture surfaces were spherical, whereas some elongated inclusions were found on CA6NM steel fracture surfaces. However, due to the very low number of these elongated particles, overall 5 elongated inclusions of about 60 mm in length among hundreds of spherical ones observed on tensile, JIc, and CVN fracture surfaces, all of the inclusions detected on CA6NM steel also could be considered as spherical. A summary of the types of inclusions detected on fracture surfaces is given in Table 9. Inclusions are categorized in three main types: Si-rich, Mn-rich, and Al-rich components. Silicon oxide is the dominant component of Si-rich type. However, remarkable amounts of manganese sulfide (MnS) also exist in this type of inclusion. The major part of Mn-rich inclusions is composed of MnS; some inclusions of this type could be considered as MnS, but portions of silicon oxide and aluminum oxide are found in most of these inclusions. An important portion of Al-rich inclusions consists of aluminum oxide and remarkable amounts of magnesium oxide and calcium oxide also detected in the inclusions of this type. Very low amounts of MnS are detected in Al-rich inclusions. Briefly, the major differences between the types of inclusions detected in CA6NM steel and 415 steel are the proportions of these three types of oxides. Silicon oxide as well as MnS are the dominant components of CA6NM steel inclusions whereas the majority of 415 steel inclusions are of aluminum oxide. SEM micrographs of Si-rich and Mn-rich complex inclusions observed in CA6NM steel as well as their EDX chemical spectrums are shown in Fig. 13(a) and (b), respectively. Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

13

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

Fig. 11. CVN fracture surfaces of CA6NM (a, c, e) and 415 (b, d, f) steels tested at 23 °C,
80 °C, and
190 °C; some intact/broken inclusions are shown by the arrows in (a) to (d).

Table 8 Inclusion size measured on the fracture surfaces of JIc and CVN tests (performed at room temperature). Material

Test

Observed surface

Number of measured inclusions

Size (mean diameter, mm)

CA6NM steel

JIc CVN

(T-S) (T-S)

40 29

12.66 ± 6.75 14.83 ± 5.89

415 steel

JIc CVN

(T-S) (T-S)

46 35

3.93 ± 2.73 4.42 ± 1.15

4. Discussion The main purpose of this section is to discuss the superior mechanical properties of the 415 steel based on the microstructural features. An increase of 75% in the carbon content of the 415 steel (Table 1) can be a reason why its strength is higher than that of the CA6NM steel since yield stress of martensite is proportional to the square root of solid solution carbon atoms [19,29]. On the other hand, prior austenite grain size is 45% finer in the 415 steel (Table 2), however, the ductility is not

Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

14

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

Fig. 12. Size distributions of inclusions measured on (a) JIc and (b) CVN fracture surfaces.

highly affected by grain refinement as reported in [19] and based on investigations by Stonsifer and Armstrong [30], this decrease results in only 3% increase in the fracture toughness. The finer microstructure in the 415 steel cannot be the main reason for 12%, 39% and, 49% improvements in its elongation at fracture, fracture toughness JIc, and CVN absorbed energy at room temperature, respectively (see Tables 5 and 7). Therefore, another microstructural feature is the origin of these improvements i.e., non-metallic inclusions. These particles differ in content, size, and nature for the steels investigated in this study. In order to shed light on how these differences affect the mechanical properties, their behavior during rupture will be discussed below.

Table 9 Types of inclusions detected in the studied materials. Material

a

Test

Inclusion type Si – rich component

Mn-rich component U U U

CA6NM steel

Tensile JIc CVN

U U U

415 steel

Tensile JIc CVN

Not detecteda

Al – rich component

U U

Due to the small size of inclusions and inclusion/dimple wall overlapping, precise detection of inclusion types was not possible.

Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

15

Fig. 13. Typical complex inclusions observed on the fracture surfaces of CA6NM steel with EDX spectrums at the right: (a) Si-rich inclusion and (b) Mn-rich inclusion.

The fracture mechanism of all the tests carried out at ambient temperature was micro-void coalescence (MVC) for both steels. However, SEM micrographs showed that the main differences between the two steels in their fracture appearances at microscopic scale are the inclusion size distribution (Figs. 6 and 12), the inclusion size (Tables 6 and 8), the dimple size, the existence of intact or broken inclusions, and the nature of inclusions (Table 9). Large inclusions of more than 10 mm in diameter were readily observed on CA6NM steel fracture surfaces whereas these inclusions were rarely detected on 415 steel fracture surfaces. The local densities of these large inclusions measured on tensile, JIc and CVN fracture surfaces were 55% (Fig. 6), 63% (Fig. 12(a)), and 79% (Fig. 12(b)), respectively, in the case of CA6NM steel, but their existence could be considered as negligible (3% or less) on 415 steel fracture surfaces, as shown in Figs. 6 and 12(a) and (b). In addition, the formation of dimples was facilitated by the rupture of inclusions [16] in CA6NM steel. Large dimples besides very fine ones existed on CA6NM fracture surfaces, whereas uniform dimple distributions were observed in the case of 415 steel. Therefore, the premature formation of micro-voids on inclusions broken during loading and on intact inclusions of more than 10 mm in diameter occurred in large number in the case of CA6NM steel. These micro-voids grew and became larger as deformation increases; at a certain deformation level, a second series of micro-voids nucleated from smaller inclusions. At this stage, large microvoids were connected together by smaller ones i.e., micro-void coalescence occurred and final rupture took place. On the other hand, in the case of 415 steel, the number of inclusions of more than 10 mm in diameter is very low and in addition, the only way for micro-void formation is the decohesion of the matrix-inclusion interface. In other words, it can be said that most inclusions in 415 steel have more or less the same chance for micro-void nucleation. As a result, the formation of microvoids occurred on most inclusions at approximately the same strain level and hence, more uniform dimples were formed and the local inclusion density may control the fracture mechanism. Briefly, the inclusion sizes given in Table 4 (measured on metallographic sections) are different from that provided in Tables 6 and 8 (measured on tensile, JIc, and CVN fracture surfaces tested at room temperature). In the case of CA6NM steel, this difference shows that large inclusions (larger than 10 mm) drive the rupture as they are the basis of dimples initiated at

Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

16

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

lower strains. On the other hand, in 415 steel, large inclusions seem not to play an essential role on void nucleation and it is rather the frequency of inclusions that drives the nucleation mechanism as numerous inclusions smaller than 10 mm in diameter are found on 415 steel fracture surfaces. The higher density of particles on fracture surfaces than on a random surface generated in the alloy suggests that the local inclusion density, and not the inclusion size, may control the fracture mechanism in 415 steel. According to the above discussion, it can be said that the high frequency of premature micro-voids formation in CA6NM steel decreased the elongation at fracture, the fracture toughness JIc, and the CVN absorbed energy at room temperature. On the contrary, the formation of micro-voids on smaller inclusions retard the fracture until higher strains are reached increasing the elongation at fracture, fracture toughness JIc and CVN values at room temperature of 415 steel. For the CVN tests performed at
80 °C (Fig. 11(c) and (d)), cleavage is dominant and few dimples were observed among cleavage facets. MVC was happening between regions that generate cleavage facets, but the amount of ductile fracture is negligible. On the other hand, at
190 °C, as shown in Fig. 8, both steels absorb the same fracture energy, and fractures occurred by pure cleavage despite of larger inclusions in CA6NM steel, as shown in Fig. 11(e) and (f). Cleavage crack propagation occurs along favorably oriented crystallographic planes [16]. Therefore, even at low temperatures, there can be some misoriented regions which cannot produce cleavage. These regions break at the final stage of rupture with dimples features, and the most important part of the energy is absorbed during the growth of these dimples [31]. However, they act for the initiation stage as stress raisers and the inclusion size does not play a significant role and this is only the shape which contributes to rupture, since very small plastic deformation is required for cleavage. Nearly the same fracture energy was observed in the cases of the tests performed at
80 °C and lower (Fig. 8), it can be concluded that different inclusion size of CA6NM and 415 steels did not affect the fracture energy. It still remains one question; what is the contribution of the different natures of inclusions detected on CA6NM and 415 steels fracture surfaces? For the tests performed at room temperature, both large inclusions of more than 10 mm in diameter and small ones (those of less than 10 mm in diameter) were found to be broken on CA6NM steel fracture surfaces. These broken inclusions were Si-rich and Mn-rich. All inclusions were Al-rich on 415 steel fracture surfaces and even large ones were intact (not broken). Therefore, it seems that the low resistance of Si-rich and Mn-rich inclusions against rupture promote micro-void formation since rupture of inclusions makes the micro-void nucleation easier [16]. Moreover, it has been shown that MnS promote micro-void nucleation [20], and as high amounts of MnS were observed in Si/Mn-rich inclusions (see Section 3.4), eventually promoting the lower resistance of Si/Mn-rich inclusions. Accordingly, it can be said that the absence of low resistant inclusions also contributed to improve the mechanical properties of the 415 steel. Finally, it can be concluded that the most effective way to improve the mechanical properties of casting 13% Cr-4% Ni martensitic stainless steels is to decrease as much as possible the volume fraction and especially the size of the nonmetallic inclusions in the microstructure. Very fine inclusions homogenously dispersed and spaced as far apart as possible will result in the highest elongation at fracture, plane strain fracture toughness, and CVN absorbed energies at room temperature. 5. Conclusions The main purpose of this paper was to characterize the effects of microstructure and inclusion characteristics on the mechanical properties of a cast and a wrought 13% Cr-4% Ni martensitic stainless steels. Tensile, fracture toughness JIc, and CVN tests were performed. The Size, surface density, volume fraction, and spacing of inclusions were determined on metallographic sections. The size and the nature of inclusions on fracture surfaces were determined using SEM-EDX. Austenite contents were measured on metallographic sections as well as below JIc fracture surfaces using XRD. The following conclusions have been drawn: – As compared to the wrought steel, the CA6NM steel has higher surface density and mean diameter of inclusions. The inclusion volume fraction is about 2.5 times higher, whereas the calculated inclusion spacing is approximately equal for both steels. – The yield stress, tensile strength, elongation at fracture, fracture toughness JIc, and CVN absorbed energy (at room temperature) of the cast steel are all lower than those of the wrought one. – For both steels, dimpled rupture occurs for the tests carried out at room temperature. The fracture modes for low temperature CVN tests are semi-cleavage and cleavage, however, cleavage is the dominant mechanism and both steels absorb the same fracture energies. – The inclusion size measured on fracture surfaces, tested at room temperature, is at least three times larger in the case of the cast steel. The majority of inclusions detected on fracture surfaces of tensile, JIc, and CVN tests are larger than 10 mm. On the contrary, almost all the inclusions on the wrought steel fracture surfaces have diameters smaller than 10 mm. – Si/Mn-rich and Al-rich complex inclusions are the two main types of defects detected on the fracture surfaces of the cast and the wrought steels, respectively. The Si/Mn-rich inclusions are ruptured, whereas Al-rich inclusions are found to be intact in the case of the wrought steel. The Si/Mn-rich inclusions have a lower resistance against rupture and against

Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002

F. Foroozmehr et al. / Engineering Fracture Mechanics xxx (2017) xxx–xxx

17

micro-void nucleation as compared to Al-rich inclusions. Ruptured Si/Mn-rich inclusions and those having diameters of larger than 10 mm are the main sites of micro-void nucleation in the cast steel. Inversely, micro-voids nucleate from intact Al-rich inclusions which nearly all of those are of less than 10 mm in diameter in the case of the wrought steel. – The higher mechanical properties of the wrought steel are originated from its lower inclusion volume fraction, more numerous inclusions having diameters lower than 10 mm, and a more resistant inclusion type (Al-rich) against rupture and against micro-void formation. – At the crack tip in a JIc specimen, the spread of the zone where the TRIP effect occurs in the wrought steel is about 6 mm in a direction perpendicular to the fracture surface. Outside this zone, the strains are not high enough to transform the austenite into martensite.

Acknowledgements This study is supported by the Natural Science and Engineering Research Council of Canada (NSERC), Hydro-Québec, and Alstom Power Hydro. SEM-EDX, XRD and Rietveld analyses were carried out at the Institut de Recherche d’Hydro-Québec (IREQ). The experimental assistance by Mrs. Manon Provencher, Mr. Étienne Dallaire, Mr. Alexandre Lapointe, technologists at IREQ, and Mrs. Josée Laviolette, technologist at École Polytechnique de Montréal, is gratefully acknowledged. References [1] Folkhard E, Rabensteiner G. Welding metallurgy of stainless steels. Wien; New York: Springer-Verlag; 1988. [2] Bilmes P, Llorente C, Ipina J Pérez. Toughness and microstructure of 13Cr4NiMo high-strength steel welds. J Mater Eng Perform 2000;9:609–15. [3] Das C, Albert S, Bhaduri A, Srinivasan G, Ramasubbu V. Effect of minor change in composition on toughness of weldmetal for repair of turbine blades made of martensitic stainless steel. Sci Technol Weld Joi 2008;13:159–66. [4] Thibault D, Bocher P, Thomas M, Lanteigne J, Hovington P, Robichaud P. Reformed austenite transformation during fatigue crack propagation of 13%Cr– 4%Ni stainless steel. Mater Sci Eng: A 2011;528:6519–26. [5] Bilmes P, Solari M, Llorente C. Characteristics and effects of austenite resulting from tempering of 13Cr-NiMo martensitic steel weld metals. Mater Charact 2001;46:285–96. [6] Webster D. Optimization of strength and toughness in two high-strength stainless steels. Metall Trans 1971;2:1857–62. [7] Huang F, Gelles D. Influence of specimen size and microstructure on the fracture toughness of a martensitic stainless steel. Eng Fract Mech 1984;19:1–20. [8] Kotecki D. Welding of stainless steels. ASM handbook, vol. 6. ASM International; 1993. p. 677–707. [9] Iwabuchi Y, Kobayashi I. A study of toughness degradation in CA6NM stainless steel. Mater Sci Forum 2010;654–656:2515–8. [10] Lippold JC, Kotecki DJ. Welding metallurgy and weldability of stainless steels. Hoboken, NJ: Wiley-Interscience; 2005. [11] Antolovich S. Alloy design for fatigue and fracture. ASM handbook, vol. 19. ASM International; 1996. p. 27–41. [12] Lee S, Majno L, Asaro R. Correlation of microstructure and fracture toughness in two 4340 steels. Metall Trans A 1985;16:1633–48. [13] Garrison Jr WM. Controlling inclusion distributions to achieve high toughness in steels. Iron Steel Technol 2007;4:132–9. [14] Choudhary P, Garrison Jr WM. The effect of inclusion type on the toughness of 4340 steel. Mater Manuf Process 2010;25:180–4. [15] Gurland J, Plateau J. The mechanism of ductile rupture of metals containing inclusions. Trans ASM 1963;56:442–54. [16] Anderson TL. Fracture mechanics: fundamentals and applications. CRC Press; 2005. [17] Eriksson K. Influence of inclusions on the fracture toughness of a SIS 2140 type tool steel. Scand J Metall 1975;4:131–9. [18] Cox T, Low JR. An investigation of the plastic fracture of AISI 4340 and 18 Nickel-200 grade maraging steels. Metall Trans 1974;5:1457–70. [19] Hertzberg RW. Deformation and fracture mechanics of engineering materials. 4th ed. New York; Toronto: Wiley; 1996. [20] Maloney JL, Garrison WM. The effect of sulfide type on the fracture behavior of HY180 steel. Acta Mater 2005;53:533–51. [21] ASTM-International. Standard specification for castings, iron-chromium, iron-chromium-nickel, corrosion resistant, for general application. In: A743/ A743M, editor. USA: ASTM International; 2010. [22] ASTM-International. Standard specification for chromium and chromium-nickel stainless steel plate, sheet, and strip for pressure vessels and for general applications. In: A240/A240M, editor. USA: ASTM-International; 2011. [23] ASTM-International. Standard test methods for tension testing of metallic materials. In: E8/E8M – 09, edtior. USA: ASTM; 2009. [24] ASTM-International. Standard test method for measurement of fracture toughness. In: E1820-11, edtior. USA: ASTM; 2011. [25] Chen J, Verreman Y, Foroozmehr F, Lanteigne J. Fracture toughness of 13Cr4NiMo high-strength steels used in hydraulic turbine runners Montreal. In: Materials science and technology (MS&T) 2013, Montreal, Quebec, Canada. p. 702–11. [26] ASTM-International. Standard test methods for notched bar impact testing of metallic materials. In: E23-07, edtior. USA: ASTM; 2007. [27] Young RA. Introduction to the Rietveld method. In: Young RA, editor. The Rietveld method. UK: Oxford University Press; 1993. p. 1–38. [28] Godin S. Effet d’un enrichissement en nickel sur la stabilité mécanique de l’austénite de réversion lorsque soumise à de la fatigue oligocyclique. École de technologie supérieure; 2014. [29] Bhadeshia HKDH, Honeycombe SR. 5 – formation of martensite. In: Honeycombe SR, Bhadeshia HKDH, editors. Steels, 3rd ed. Oxford: ButterworthHeinemann; 2006. p. 95–128. [30] Stonesifer FR, Armstrong RW. The physical metallurgy of fracture. In: Taplin DMR, editor. Advances in research on the strength and fracture of materials, vol. 2A. Pergamon Press; 1977. p. 1–6. [31] Campbell FC. Fatigue and fracture: understanding the basics. ASM International; 2012.

Please cite this article in press as: Foroozmehr F et al. Effect of inclusions on fracture behavior of cast and wrought 13% Cr-4% Ni martensitic stainless steels. Engng Fract Mech (2017), http://dx.doi.org/10.1016/j.engfracmech.2017.02.002