1 Introduction

Pyrowear 53 steel is a carburizing grade of steel developed in the 1970s, which possesses high hardenability, good temper resistance, and an excellent combination of high strength and fracture toughness [1, 2]. Most of all, it shows high hot hardness and the ability to retain its strength under limited lubrication owing to a high content of molybdenum and copper. The addition of molybdenum improves its heat and wear resistance [2, 3], while copper enhances its hardness and strength by forming numerous fine Cu-based precipitates during tempering [4]. Due to its increased beneficial properties and safety margin in comparison to other commonly used carburized steels (e.g., 9310 steel), Pyrowear 53 steel is utilized in aviation and military applications where performance and safety overcome cost considerations, especially in power transmission components, such as gear wheels or shafts [1, 2]. Its typical heat treatment process involves [2, 5]: (1) carburizing at a temperature of 870–930 °C, (2) oil or gas quenching at a temperature of 905–920 °C, (3) subzero treatment at about −75 °C, and (4) tempering at 150–290 °C. As a result, such heat-treated elements made of Pyrowear 53 steel exhibit a high hardness and wear resistance in the carburized layer while their core retains high fracture toughness and the ability to carry large impact loads [1, 2]. However, the material becomes inhomogeneous due to the higher carbon content in the carburized layer, which leads to a need to investigate its local properties in order to understand its overall mechanical performance and fracture behavior.

The small punch test (SPT) is one of the methods that enables the characterization of local mechanical properties of small-scale materials with a limited volume [6, 7], thin layers [8], materials extracted from in-service components [9], or weldments [10]. During the SPT, a disc-shaped or rectangular specimen is pushed through a die using a small hemispherical punch, while load–displacement data is recorded and adopted for the estimation of the tensile or fracture properties [6, 11]. It has been widely used in recent years for yield strength and ultimate tensile strength estimation [11, 12], along with the evaluation of creep properties [13, 14], ductile-to-brittle transition temperatures [15, 16], and hydrogen embrittlement [17, 18]. Nevertheless, the most sophisticated application of SPT is the determination of fracture toughness. The first attempts were mostly related to the correlation of the equivalent fracture strain or the total energy under load–deflection curves from the SPT with a plane strain, JIC, integral value [19,20,21,22]. It has been shown that the biaxial fracture strain was roughly linear with respect to the JIC integral for various ferrous materials [19] and aluminum alloys [20]. In turn, Ha and Fleury [21] linked the JIC fracture toughness with the total energy under the SPT curve by subtracting the plastic deformation energy, which was calculated based on the microscopic observations of deformed SPT samples. Wang et al. [22] followed the same energy-based approach, which was additionally supported by the finite element analysis of SPT sample deformation.

In the following years, many research groups have introduced a crack in SPT specimens as a stress concentrator. Various types of notches have been investigated, e.g., a U-shaped surface longitudinal notch [23, 24], a surface circular notch [25, 26], or a through-thickness lateral notch [27, 28]. The proposed methodologies for these pre-notched SPT specimens can be divided into the following two approaches: (1) the estimation of the fracture energy under the load–displacement curve [25, 26] or (2) the use of the crack tip opening displacement (CTOD) concept [23, 24, 28]. The latter approach requires a precise measurement of notch opening during the tests by an endoscope inspection camera [28] or performing several tests interrupted at various punch displacements in order to measure notch opening by additional microscope observations [23, 24]. The energy-based approach seems to be more promising but needs further improvement to correlate the SPT results with standard JIC tests [28].

The aim of the current study was to investigate the local mechanical properties (especially fracture toughness) of different zones in carburized Pyrowear 53 steel using SPT. Firstly, the Pyrowear 53 steel was heat treated (quenched and tempered) at different processing parameters to obtain various results for the microstructure and fracture toughness of the core material. It was characterized in terms of its microstructure, tensile properties, and fracture toughness using standardized test specimens. The results obtained for the core material in the standardized fracture toughness tests were correlated with the SPT results. SPT samples with U-shaped longitudinal notches were used since this specimen type is frequently employed in the literature and is relatively easy to prepare by electric discharge machining (EDM). The energy-based approach was utilized to estimate the fracture toughness using the SPT method. In the second part of this study, Pyrowear 53 steel was carburized at 925 °C and divided into the following zones: (1) a carburized layer (up to 0.5 mm from the surface), (2) a transition layer (from 0.5 to 1.5 mm), and (3) a core zone (more than 1.5 mm). Each zone was characterized in terms of its microstructure and tensile properties using miniaturized test specimens. Finally, the fracture toughness of the carburized and transition layers was determined based on the obtained correlation of the SPT results with the standardized fracture toughness tests.

2 Material and methods

2.1 Material preparation

The chemical composition of the investigated Pyrowear 53 steel is presented in Table 1. It was supplied in the form of rods with a diameter of 7 inches which were cut into cylindrical slabs having a height of 30 mm. These samples were then austenitized at a temperature of 910 or 925 °C and quenched by nitrogen gas flowing at a pressure of at least 9 bar (higher cooling rate) or below 9 bar (lower cooling rate). After quenching, a subzero treatment below −75 °C and two-step tempering at 230 °C were performed. These heat-treated samples were named QT_1, QT_2, and QT_3 as summarized in Table 2. These products, representing a core of gears made of Pyrowear 53 steel, were machined into specimens for fracture toughness, uniaxial tensile, and small punch tests as presented in Fig. 1a. Additionally, rectangular samples with dimensions of 5 mm × 10 mm × 45 mm were cut from the cylindrical slabs (Fig. 1a) and then subjected to low pressure carburizing at 925 °C (as shown in detail in [3]), followed by the same processing procedure mentioned above, i.e., austenitizing at 910 or 925 °C, quenching by high pressure nitrogen gas, subzero treatment, and two-step tempering. These rectangular specimens, named CQT_1, CQT_2, and CQT_3 (Table 2), were used for the characterization of carburized layers in Pyrowear 53 steel.

Table 1 Chemical composition of Pyrowear 53 steel measured by Spark-EOS [29]
Full size table
Table 2 Gas quenching parameters applied to Pyrowear 53 steel
Full size table
Fig. 1
figure 1

a The location of test specimens in the cylindrical slabs of Pyrowear 53 steel and technical drawings of: b CT fracture toughness sample, c standard tensile sample, and d miniaturized tensile sample

Full size image

2.2 Fracture toughness, tensile properties and hardness characterization

The fracture toughness of Pyrowear 53 steel after quenching and tempering was determined according to the ASTM E399 standard using compact tension (CT) specimens with a thickness of B = 25 mm (Fig. 1b) and an MTS Landmark servo-hydraulic testing machine equipped with a 250 kN load cell and an MTS 632.30F-30 OPT.005 clip-on gage. Each material condition was represented by five samples. Sinusoidal loading was applied with a stress ratio of 0.1 at a frequency of 20 Hz to obtain a precrack with a final length of 26 mm, while making sure that the maximum stress intensity factor, Kmax, at the final stage of precracking did not exceed 30 MPa m1/2. After precracking, the final loading to fracture was conducted at a loading rate of 1 kN/s. Uniaxial tensile tests of the Pyrowear 53 steel after quenching and tempering were performed according to the ISO 6892–1 standard on an MTS 810 servo-hydraulic testing machine using cylindrical test specimens with a diameter of 5 mm and gauge length of 25 mm (Fig. 1c). The tensile properties of the carburized layer were assessed by uniaxial tensile tests on miniaturized flat samples with a cross section of 0.6 mm × 0.8 mm and a gauge length of 4 mm (as shown in Fig. 1d) using a ZwickRoell Z005 testing machine with a 5 kN loading capacity and a digital image correlation (DIC) system for strain measurements (described in more detail in [30, 31]). All the tensile tests were performed at an initial strain rate of 10−3 s−1. The tensile test results were utilized to evaluate the 0.2% offset yield strength (YS), the ultimate tensile strength (UTS), the elongation to failure (A5), and the strain hardening exponent (n), which was calculated based on the power law describing the relation of true stress (σ), true strain (ε), and strength coefficient (K) within a uniform elongation regimen expressed as follows [32]:

$$\sigma =K{\varepsilon }^{n}$$
(1)

The strain hardening exponent was evaluated for a strain range of 2–4%. Each material condition was represented by three tensile test specimens. The measurement uncertainties were calculated according to Annex J from the ISO 6892–1 standard as the combined uncertainty of the standard deviations of the obtained results from three different measurements and the accuracy of the measuring equipment, i.e., the accuracy of the load cell (± 0.5%), extensometer (± 0.5%) or DIC strain measurement system (± 1%), gauge length, and cross-section measurements (± 1%). Finally, Vickers hardness tests were conducted at a load of 500 g using an Innovatest Falcon 500 hardness tester. The measurement uncertainties were estimated based on the standard deviations and the equipment errors (i.e., the accuracy of the load cell ± 1% and diagonal length measurement ± 2%) according to the ISO/IEC Guide 98–3.

2.3 Microstructural characterization

To obtain the overall material characteristics, the microstructure of Pyrowear 53 steel was characterized using a Hitachi SU8000 scanning electron microscope (SEM). Sections for microscopic examination were ground, polished, and etched with a solution of 2% nital. Additionally, the Pyrowear 53 steel, after quenching and tempering, was investigated by electron backscatter diffraction (EBSD) using a Hitachi SU-70 SEM equipped with a Schottky emitter and a Bruker e-Flash HD detector. EBSD samples were polished with an argon ion beam using a Hitachi IM4000 Ion Milling system. EBSD analysis was conducted on an area of around 3000 µm2 using a step size of 0.1 µm, and the obtained data were used for the estimation of the volume fraction of the retained austenite, martensite, and bainite.

2.4 Small punch tests

The experimental setup used for SPTs is presented in Fig. 2a-c. It consisted of upper and lower dies with a receiving hole diameter of rp = 5.4 mm and a spherically ended punch having a radius of 1.25 mm. SPT samples (Fig. 2d) having a diameter of d = 8 mm, a thickness of t = 0.55 mm, and a U-shaped longitudinal notch with a width of w = 0.27 mm and a depth of about h = 0.25 mm were obtained by EDM. Before the tests, SPT specimens were ground using 1200 grit SiC paper to remove the oxidized layer created during the EDM process and obtain a final thickness of 0.500 ± 0.005 mm. The SPT samples were mounted between the lower and upper dies and clamped by a mounting screw with a 10 Nm torque. The SPTs were performed using a ZwickRoell Z005 testing machine at a crosshead speed of 2 mm/min until the load had decreased by 20%. The deflection of the SPT specimens was measured by an MTS 634-12F-25 extensometer attached to the lower surface of the sample. After SPT, each sample was observed using a Hitachi SU8000 SEM to measure the crack length and the degree of sample deformation. Each material condition was represented by five SPT samples, while the measurement uncertainties were estimated based on the standard deviations and the equipment errors (i.e. the accuracy of load cell ± 0.5% and extensometer ± 0.5%).

Fig. 2
figure 2

a-c The experimental setup and d test specimens used for SPTs: (a) a close-up of the specimen mounting system, (b) technical drawing of the SPT chamber, and (c) actual image of the SPT setup

Full size image

3 Results and discussion

3.1 Microstructure and mechanical properties of the core

Figure 3 presents SEM images and EBSD band contrast maps of the Pyrowear 53 steel after quenching and tempering with different processing parameters (i.e., QT_1, QT_2, and QT_3). All the samples consisted of lath martensite, bainite, and numerous carbide particles containing Mo, Fe, Cr, V, and Si, which are typical microstructural constituents for heat-treated Pyrowear 53 steel [5, 29]. Due to the differences in the crystal lattice structure, the EBSD analysis also showed the presence of residual austenite at levels of 1.4, 1.5, and 2.3% for QT_1, QT_2, and QT_3, respectively. The film-like austenite was located between the bainite and martensite laths, as marked by the red color in the band contrast maps. In order to differentiate martensite and bainite, the band contrast maps from the EBSD measurements were analyzed according to the procedure of Baek et al. [33] and Wang et al. [34]. In general, the martensite shows greater lattice imperfections as compared to bainite, so the quality and intensity of its Kikuchi pattern are lower, as is the band contrast value. Based on the obtained distribution of the band contrast value, a reference point for differentiating martensite and bainite was defined as 128, i.e., the lower band contrast value corresponded to martensite while the higher represented bainite. The same reference point was set previously by Baek et al. [33] and Zisman et al. [35]. The analysis of the band contrast distributions showed that the fraction of martensite in the Pyrowear 53 steel after quenching and tempering gradually increased in the following order: 40.7% for QT_1 < 47.2% for QT_2 < 52.8% for QT_3. In turn, the bainite content decreased from 57.9% for QT_1 to 51.3% for QT_2 and 44.9% for QT_3.

Fig. 3
figure 3

Microstructure of the Pyrowear 53 steel after quenching and tempering using different processing parameters shown by SEM images, EBSD band contrast maps, and band contrast value distributions. The red color in the band contrast maps represents the retained austenite

Full size image

The main aim of changing the microstructure of Pyrowear steel by varied quenching and tempering protocols was to obtain variation in the fracture toughness of the core material. All mechanical properties from the fracture toughness, uniaxial tensile, and hardness tests are summarized in Table 3, while the representative tensile stress–strain curves and load-crack opening displacement (COD) curves from the fracture toughness tests are presented in Fig. 4. All the samples exhibited similar tensile properties, i.e., YS of 892, 899, and 910 MPa, UTS of 1123, 1130, and 1148 MPa, and A5 = 16.1, 16.5, and 16.2% for QT_1, QT_2, and QT_3, respectively. The observed slight increase in YS (from 892 MPa for QT_1 to 910 MPa for QT_3) was in good agreement with the results of the hardness measurements, i.e., the Vickers hardness for QT_1, QT_2, and QT_3 was 374, 378, and 388 HV0.5, respectively. Most of all, the fracture toughness of the Pyrowear 53 after the various quenching and tempering processes was differentiated significantly. The calculated fracture toughness, KQ, gradually increased in the following order: 102 ± 9 MPa m1/2 for QT_1 < 118 ± 9 MPa m1/2 for QT_2 < 134 ± 4 MPa m1/2 for QT_3. These differences may result from the samples’ different phase compositions, especially the austenite and martensite content. As shown in Fig. 3, the volume fraction of retained austenite and martensite grew in the same order, i.e., QT_1 < QT_2 < QT_3. Prasad and Putatunda [36] showed for high carbon steel that the fracture toughness can be significantly improved by raising the austenite volume fraction from 12 to 25%. Similar findings were observed by Wu et al. [37] for a low carbon TRIP steel with a martensite-bainite microstructure and a lower content of retained austenite ranging from 1 to 9.2%. These results clearly show that retained austenite, as a more ductile phase, is highly effective in improving the fracture toughness of high-strength steels by stress-relief or crack-arrest effects [38], as well as by martensitic transformation ahead of a crack tip at high austenite content [37]. At the relatively low fraction of retained austenite in the current study (i.e., 1.4–2.3%), the enhancement of fracture toughness is more associated with the co-deformation of film-like austenite and tempered martensite laths, which leads to a higher resistance to crack propagation with increasing austenite content, as also shown by Hu et al. [39] for a low-carbon medium-manganese high-strength steel after quenching and tempering. The increase in fracture toughness can also be linked with an increasing content of tempered martensite. Tomita [40] and Abdollah-Zadeh et al. [41] showed for medium carbon steels that the tempered martensite microstructure may provide higher fracture toughness than bainitic or mixed martensitic-bainitic microstructures. In fact, Zhang and Knott [42] clearly proved this to be the case for a low-carbon, low-alloy steel by achieving the following fracture toughness: 45.8 MPa m1/2 for a fully bainitic microstructure; 54.6 MPa m1/2 for a mixed 45% bainite + 55% martensite microstructure; 61.1 MPa m1/2 for a mixed 30% bainite + 70% martensite microstructure; and 92.0 MPa m1/2 for a fully tempered martensite. Our results and these other studies allow us to conclude that the increase in the fracture toughness of QT_1, QT_2, and QT_3 samples results from the increasing content of retained austenite and tempered martensite (as shown in Fig. 3).

Table 3 Mechanical properties of the Pyrowear 53 steel after varied quenching and tempering processes
Full size table
Fig. 4
figure 4

a The representative stress–strain curves from uniaxial tensile tests and b load-crack opening displacement (COD) curves from fracture toughness tests of the Pyrowear 53 steel after varied quenching and tempering processes

Full size image

3.2 SPT results for the core

In the next step, QT_1, QT_2, and QT_3 samples showing fracture toughness values of 102 ± 9, 118 ± 9, and 134 ± 4 MPa m1/2, respectively, were tested by the SPT method. The representative load–deflection curves are shown in Fig. 5a, while Table 4 summarizes the values of the maximum force (Fm), deflection at fracture (uf), and the total energy under the load–deflection curve (WT) determined up to a 20% load drop after reaching the maximum force according to the CWA standard [43]. The obtained load–deflection curves exhibited four deformation stages, which are typically observed during SPT of a ductile material [22, 44]: (I) elastic bending, (II) plastic bending, (III) membrane stretching, and (IV) plastic instability (as schematically presented in Fig. 5a). The maximum force, Fm, increased in the following order: 1080 ± 22 N for QT_1, 1192 ± 24 N for QT_2, and 1208 ± 16 N for QT_3, while the deflection at fracture, uf, was 1.21 ± 0.01, 1.26 ± 0.01, and 1.24 ± 0.01 mm for QT_1, QT_2, and QT_3, respectively. The WT values calculated based on the SPT load–deflection curves were as follows: 607 ± 34 mJ for QT_1, 667 ± 29 mJ for QT_2, and 700 ± 16 mJ for QT_3, showing a similar tendency to the fracture toughness determined using the standardized CT specimens (Table 2).

Fig. 5
figure 5

a The representative load–deflection curves and b correlation of fracture energy integral, JSPT, values from SPT with JQ values obtained based on the standardized fracture toughness tests of the Pyrowear 53 steel after varied quenching and tempering processes

Full size image
Table 4 SPT results for Pyrowear 53 steel after varied quenching and tempering processes
Full size table

Since the deformation of SPT samples covers elastic–plastic bending, membrane stretching, and the fracture mechanism, the calculated total energy under the load–deflection curves (WT) can be expressed as [21]:

$${W}_{T}={W}_{p}+{W}_{f}$$
(2)

where Wp and Wf are the energy of elastic–plastic deformation and fracture energy, respectively. For the expression of the elastic–plastic deformation energy of a thin sheet deformed to a hemispherical shape, Atkins and Mai [45] have proposed the following formula:

$${W}_{p}=\left(\pi {{r}_{p}}^{2}t\right)\frac{{\sigma }_{0}{{\varepsilon }_{p}}^{n+1}}{n+1}$$
(3)

where rp is the radius of the lower die, t is the specimen thickness, σ0 is the yield strength, εp is the average plastic strain, and n is the strain hardening exponent. Taking into account that SPT samples deform non-uniformly, Ha and Fleury [21] divided the plastic deformation energy for the part related to the deformation of the specimen section under the punch (inside part of the necking) and the part outside of the necking area. SPT samples investigated within the current study possessed a U-shaped notch that limited their uniform deformation, and thus the plastic deformation energy was split into a part, W1, corresponding to the deformation of the specimen section under the notch (marked in brown in Fig. 6a), which underwent more pronounced thinning during SPT, and a second part, W2, related to the more uniform deformation of the remaining section of the SPT specimen (marked in gray in Fig. 6a). The following expressions were used for the calculation of the plastic deformation energy of each zone:

$${W}_{1}=\left[{r}_{p}w\left(t-h\right)\right]\frac{{\sigma }_{0}{{\varepsilon }_{1}}^{n+1}}{n+1}$$
(4)
$${W}_{2}=\left[\left(\pi {{r}_{p}}^{2}-{r}_{p}w\right)t\right]\frac{{\sigma }_{0}{{\varepsilon }_{2}}^{n+1}}{n+1}$$
(5)

where rp = 5.4 mm is the radius of the lower die, w = 0.27 mm is the notch width, t is the specimen thickness, h is the notch depth, and σ0 and n are the yield strength and strain hardening exponent, respectively, as summarized in Table 3. ε1 and ε2 are the average plastic deformation of the specimen section under the notch and in the remaining part, respectively, which were estimated based on SEM observations from the following formulas:

$${\varepsilon }_{1}=\text{ln}\left(\frac{t-h}{{t}_{1}}\right)$$
(6)
$${\varepsilon }_{2}=\text{ln}\left(\frac{t}{{t}_{2}}\right)$$
(7)

where t1 and t2 are the average specimen thickness after thinning during SPT in the specimen section under the notch and in the remaining part, respectively, as presented schematically in Fig. 6a–b.

Fig. 6
figure 6

SEM images of the SPT sample cut from Pyrowear 53 steel after quenching and tempering: a cross section along the U-shaped notch, b cross section transverse to the U-shaped notch, c top view, d fracture surface with e dimple size distribution

Full size image

All the calculated plastic deformation energy and fracture energy values are summarized in Table 4. It clearly shows that the majority of the total energy was consumed by the plastic deformation of the specimen in the area outside the U-shaped notch, i.e., W2 = 426–466 mJ. The plastic energy required for the deformation of the specimen section under the notch was in the range of W1 = 48–52 mJ. The remaining energy spent on the fracture mechanism increased gradually from Wf = 133 ± 25 mJ for QT_1, to 149 ± 22 mJ for QT_2, and to 210 ± 10 mJ for QT_3. In the next step, the estimated Wf values were recalculated to fracture energy integral (JSPT) values using the following formula:

$${J}_{SPT}= \frac{{W}_{f}}{{a}_{1}\cdot \left(t-h\right)}$$
(8)

where a1 is a crack length that propagated along the U-shaped notch, as shown in Fig. 6c. In this manner, the calculated fracture energy integrals, JSPT, were assumed to be the work spent on the fracture (Wf) of the cracked area under the notch, which is similar to the approach taken earlier by Ha and Fleury [21]. The calculated JSPT integral grew in the following order: 84 ± 17 kJ/m2 for QT_1 < 103 ± 15 kJ/m2 for QT_2 < 139 ± 8 kJ/m2 QT_3, showing the same tendency as the fracture toughness of the standardized CT specimens (Table 3). In order to correlate the SPT results with the standardized tests, the fracture toughness KQ of CT specimens was recalculated to the JQ integral value from the following equation [46]:

$${J}_{Q}=\frac{{{K}_{Q}}^{2}\left(1-{\nu }^{2}\right)}{E}$$
(9)

where ν = 0.3 is the Poisson ratio, and E is Young’s modulus. The estimated JQ integrals of 46.2 ± 8.1 kJ/m2 for QT_1, 62.0 ± 7.4 kJ/m2 for QT_2, and 80.0 ± 4.3 kJ/m2 for QT_3 were plotted with the corresponding JSPT integrals as shown in Fig. 5b, allowing the following formula to be determined, which correlates the JSPT values with the JQ integrals obtained based on the standard fracture toughness tests:

$${J}_{Q}={0.575J}_{SPT}+0.305$$
(10)

This formula is used for the determination of the fracture toughness of the carburized layer in the following sections.

3.3 Microstructural and mechanical properties of the carburized layer

Firstly, hardness distribution profiles were obtained for Pyrowear 53 steel after carburizing at 925 °C followed by the same quenching and tempering processes as samples QT_1, QT_2, and QT_3. The samples investigated at this step were named CQT_1, CQT_2, and CQT_3. Figure 7 presents the obtained hardness profiles, which started with a plateau at a level of about 750 HV0.5. At a distance of about 0.5 mm from the sample surface, a gradual decrease in hardness was observed, until reaching a hardness of about 400 HV0.5 at a distance of 1.5 mm. Similar hardness profiles of carburized Pyrowear 53 steel were obtained by Filip et al. [29] and Wojtyczka and Iżowski [3]. Based on the hardness distribution profiles, the following zones were distinguished: (1) the carburized layer (up to 0.5 mm from the surface), (2) the transition layer (from 0.5 to 1.5 mm), and (3) the core zone (more than 1.5 mm). The microstructure of each zone was then characterized by SEM as presented in Fig. 8. The carburized layer was composed of fine plates of high-carbon martensite and numerous carbide particles (Fig. 8a). In the transition layer, the fraction of high-carbon martensite lowered and lath martensite and bainite started to appear (Fig. 8b), while the core material comprised packets of lath martensite, bainite, and fine carbide particles (Fig. 8c), i.e., the same microstructure as the samples after the quenching and tempering processes (Fig. 3).

Fig. 7
figure 7

Hardness distribution profiles for Pyrowear 53 steel after carburizing at 925 °C followed by varied quenching and tempering processes. The red dotted lines show the locations of the SPT and miniaturized tensile test specimens

Full size image
Fig. 8
figure 8

Microstructure of Pyrowear 53 steel after carburizing at 925 °C followed by quenching and tempering in the a carburized layer, b transition layer, and c core

Full size image

To determine the tensile properties of each zone of the Pyrowear 53 samples after carburizing, miniaturized tensile samples (Fig. 1d) were cut in the locations marked by red dotted lines in Fig. 7. A comparison of the stress–strain curves for each zone is presented in Fig. 9a, while their mechanical properties are summarized in Table 5. The core material was characterized by YS = 933–976 MPa, UTS = 1125–1166 MPa, and A5 = 13.0–14.6%, depending on the applied processing parameters. These strength values were slightly higher than those obtained for the standard cylindrical test specimens presented earlier in Table 3 (i.e., YS = 892–910 MPa, UTS = 1123–1148 MPa). These discrepancies in YS and UTS values are not particularly affected by specimen size effects. It is commonly accepted that to ensure a polycrystalline bulk behavior, the specimen thickness should be at least 6–10 times greater than the grain size [47]. Most of the grains in the core zone were in the range of 5 to 10 microns (Fig. 3), which gives about 100 grains per thickness of the miniaturized tensile samples used in this study (Fig. 1d) and minimizes the specimen size effects. Instead, the various strength values may result from the different heat capacities of the heat-treated samples of different sizes. The carburized samples (CQT_1, CQT_2, and CQT_3) with dimensions of 5 mm × 10 mm × 45 mm were more strengthened in the core zone by the analogous heat treatment than the much bigger cylindrical slabs of 7 inches diameter after quenching and tempering (QT_1, QT_2, and QT_3). These results are also consistent with the hardness measurements showing hardness values of about 374–388 HV0.5 for QT_1-QT_3 (Table 3) and 385–400 HV0.5 in the core zone of CQT_1-CQT_3 samples (Fig. 7). The higher strengthening level of the core material in the carburized samples also led to their lower strain hardening exponent (n = 0.061–0.062) as compared to the previously discussed QT_1, QT_2, and QT_3 samples (n = 0.079–0.081). Most of all, the transition and carburized layers showed improved strength, as might be expected from their hardness distribution profiles (Fig. 7). The transition layer exhibited higher strength (YS = 1061–1098 MPa, UTS = 1276–1307 MPa), but lower ductility (A5 = 12.2–12.3%), than the core material. Further enhancement of strength was observed in the carburized layer, i.e., the YS and UTS reached values of 1649–1765 MPa and 2032–2192 MPa, respectively. In turn, the elongation to failure decreased to A5 = 3.8–4.4%. It is also worth noting that the strain hardening coefficient n continuously increased from the core (n = 0.061–0.062), through the transition layer (n = 0.064–0.065) to the carburized layer (n = 0.105–0.124), which is consistent with the study of Xu et al. [48], who also observed a gradual increase in n value from 0.09 for the core material up to about 0.3 for the sub-surface area in a low carbon steel after carburizing.

Fig. 9
figure 9

a The representative stress–strain curves from uniaxial tensile tests and b load–deflection curves from SPT for the three zones (core, transition, and carburized layers) in Pyrowear 53 steel after carburizing at 925 °C followed by quenching and tempering

Full size image
Table 5 Mechanical properties of the three zones in Pyrowear 53 steel after carburizing at 925 °C followed by varied quenching and tempering processes
Full size table

3.4 SPT results for the carburized layer

To determine the fracture toughness of the different zones in the Pyrowear 53 steel after carburizing, SPT samples were also prepared from locations marked by red dotted lines in Fig. 7. The load–deflection curves representative for each zone are shown in Fig. 9b, while Table 6 summarizes the values of the maximum force (Fm), deflection at fracture (uf), total energy under the load–deflection curve (WT), calculated plastic deformation energies (W1, W2), fracture energy (Wf), and JSPT integral values. All the deformation stages typical for ductile materials were distinguished in the load–deflection curves for the core material and transition layer. The maximum force, Fm, for the core material from the carburized samples still increased in the following order: 1193 ± 21 N for CQT_1, 1271 ± 41 N for CQT_2, and 1305 ± 59 N for CQT_3, but its value was higher compared to its quenched and tempered counterparts (Table 4). This change seems to be associated with their increased YS and UTS, as presented in Table 5. The WT values calculated based on the SPT load–deflection curves were also slightly improved to: 667 ± 20 mJ for CQT_1, 727 ± 31 mJ for CQT_2, and 750 ± 24 mJ for CQT_3 (as compared to 607–700 mJ for the QT samples). However, the enhanced YS and lower strain hardening exponent, n, led to a higher plastic energy being required for the deformation of the specimen section under the notch (W1 = 54–59 mJ) and in the remaining part of the specimen (W2 = 502–545 mJ). As a result, the fracture energies, Wf, were equal to 111 ± 4 mJ for CQT_1, 125 ± 16 mJ for CQT_2, and 147 ± 14 mJ for CQT_3, which, in turn, are lower than values obtained for QT_1 (133 mJ), QT_2 (149 mJ), and QT_3 (210 mJ). The JSPT integral values calculated using the formula (8) were also decreased to 78 ± 2, 87 ± 11, and 102 ± 10 kJ/m2 for CQT_1, CQT_2, and CQT_3, respectively, which indicates the lower fracture toughness of the core of the carburized samples in comparison to the core material in the QT_1, QT_2, and QT_3. Nevertheless, the fracture toughness of carburized samples still increased in the same manner, i.e., CQT_1 < CQT_2 < CQT_3.

Table 6 SPT results for Pyrowear 53 steel after carburizing at 925 °C followed by varied quenching and tempering processes
Full size table

The load–deflection curves of SPT samples from the transition layer exhibited similar deformation stages as those from the core material, but the observed maximum forces (Fm = 1343–1361 N) were significantly higher, and the deflections at fracture uf were lower (uf = 1.19–1.20 mm). As a result, the total energy values (WT = 714–719 mJ) were at a comparable level to the core material (WT = 667–750 mJ). Due to the enhanced YS in the transition layer (i.e., 1061–1098 MPa), the plastic deformation energies W1 and W2 were increased up to 58–60 mJ and 540–557 mJ, respectively. Thus, the remaining energy Wf consumed by the fracture mechanism was equal to 102 ± 31 mJ for CQT_1, 110 ± 26 mJ for CQT_2, and 115 ± 19 mJ for CQT_3, and the JSPT integral calculated using formula (8) was as follows: 71 ± 21, 76 ± 18, and 80 ± 13 kJ/m2 for CQT_1, CQT_2, and CQT_3, respectively. Despite the similar total energy, WT, for the core and transition layer, the fracture toughness of the SPT samples from the transition layer turned out to be lower, which resulted from the higher amount of energy consumed on its plastic deformation.

In contrast to the core and transition layer, the SPT samples cut from the carburized layer cracked at the onset of the plastic bending regime (Fig. 9b). Therefore, the maximum force (Fm = 330–352 N), deflection at fracture (uf = 0.27–0.29 mm) and total energy (WT = 48–54 mJ) were significantly lower compared to the core and transition layer. SEM observations revealed that the SPT samples from the carburized layer were slightly deformed at the section under the notch (Fig. 10a), while the thinning of the remaining part of the sample was negligible. Thus, it was assumed in the calculations that W2 = 0. The plastic deformation energy, W1, reached a relatively low value of about 9 mJ. In the case of the carburized layer, the majority of the total energy under the SPT curve was consumed by the fracture mechanism (Wf = 39–44 mJ). The SPT samples representing the core material and transition layer fractured in a ductile manner along the U-shaped notch (Fig. 6d). The characteristic dimples showed a bimodal size distribution (Fig. 6e), i.e., a fraction of dimples having a size of a few microns corresponding to the size of martensite and bainite packets, and a fraction of smaller dimples (< 1 µm) resulting most likely from the presence of film-like austenite located between bainite and martensite laths (as presented in the band contrast maps in Fig. 3). In turn, the SPT specimens from the carburized layer exhibited features of brittle cracking (Fig. 10c). The size of the fine cleavage facets was mostly in the range of 0.5–1.0 µm (Fig. 10d), which corresponded to fine packets of the high-carbon martensite as shown in Fig. 8a. In particular, in addition to the crack propagating along the notch, a second transverse crack was also observed in the SPT samples from the carburized layer (Fig. 10b). Therefore, the JSPT integral value for the carburized layer was calculated from the formula:

$${J}_{SPT}= \frac{{W}_{f}}{{a}_{1}\cdot \left(t-h\right)+{a}_{2}\cdot t}$$
(11)

where a1 and a2 are the lengths of cracks that propagate along the U-shaped notch and in the transverse direction, respectively, as shown in Fig. 10b. The calculated JSPT integral for the SPT samples from the carburized layer was in the range of 8.1–9.1 kJ/m2, which is one order of magnitude lower than for the core and transition layer. According to formula (10) determined in Sect. 3.2, the JSPT integral of 8.1–9.1 kJ/m2 corresponds to the JQ integral of 5.0–5.6 kJ/m2, which gives a fracture toughness, KQ, value of 33–35 MPa m1/2 using Eq. (9). Sandor and Ferreira [49] reported a plane strain fracture toughness of the carburized layer in a low carbon SAE 5115 steel of 22 MPa m1/2. However, the fracture toughness of the core in this SAE 5115 steel was about 71 MPa m1/2. In the current study, the fracture toughness of the core in the Pyrowear 53 steel was at a level of KQ = 102–134 MPa m1/2 (Table 3), maybe implying a greater toughness of the carburized layer as well. Most of all, the obtained fracture toughness of the carburized layer of about 33–35 MPa m1/2 is in a reasonable range, while the proposed SPT methodology seems to be a prospective way of determining the fracture toughness of relatively thin layers or local changes in the fracture behavior of surface-treated materials. It may be useful for the design process of engineering components such as gear wheels. The main failure modes of gears are due to fatigue (gear tooth bending fatigue or surface contact fatigue) or impact fracture (overload tooth fracture) [50]. A better knowledge of fracture toughness in the subsurface area of gear teeth (i.e., the carburized layer) may contribute to a proper selection of heat treatment parameters or safety margins and the prevention of crack initiation leading to a terminal damage of gear wheels.

Fig. 10
figure 10

SEM images of the SPT sample cut from the carburized layer in Pyrowear 53 steel after carburizing at 925 °C followed by varied quenching and tempering processes: a cross section along the U-shaped notch, b top view, c fracture surface with d cleavage facet size distribution

Full size image

4 Conclusions

The aim of the current study was to determine the fracture toughness of different zones in carburized Pyrowear 53 steel using the SPT technique. The main conclusions are as follows:

  1. 1.

    Varying the heat treatment parameters of Pyrowear 53 steel allowed various fracture toughness values of the core material (KQ = 102, 118, and 134 MPa m1/2) measured by standardized CT specimens to be obtained. The gradual increase in the fracture toughness resulted from the increasing content of retained austenite and tempered martensite.

  2. 2.

    The SPT results of the core material showed a similar tendency to the fracture toughness tests using the standardized CT specimens. A comparison of these results enables the following formula to be determined, which correlates the fracture energy integral, JSPT, from the SPT with JQ integrals obtained based on the standard fracture toughness tests:

    $${J}_{Q}={0.575J}_{SPT}+0.305$$
  3. 3.

    Three zones were distinguished in the Pyrowear 53 steel after carburizing, i.e., the carburized layer (up to 0.5 mm from the surface), the transition layer (from 0.5 to 1.5 mm), and the core zone (more than 1.5 mm). Each zone was characterized in terms of its tensile properties using miniaturized test specimens and fracture toughness using SPT samples with U-shaped longitudinal notches. The proposed energy-based approach showed that the fracture toughness decreased from JSPT = 78–102 kJ/m2 for the core zone, through JSPT = 71–80 kJ/m2 for the transition layer, down to JSPT = 8.1–9.1 kJ/m2 for the carburized layer.