Low-Alloyed Spring Steel: Nanostructure and Strength After Austempering

Abstract

Carbide-free bainite microstructures were developed in 0.6 wt.%–2 wt.% Si spring steel via austempering at 250 °C. Heat treatment included austenization with subsequent isothermal holding at 250 °C, varying the holding duration to 1.0, 1.5, or 2.0 h with final cooling in water. X-ray diffraction, SEM investigation, tensile test, and hardness measurement were employed to study the microstructure and phase compositions of the samples. It was found that nanostructured bainite developed in the experimental steel. The distribution of distances between phase borders was determined via digital processing of SEM micrographs, and the mode distance was found to be 30 nm. The analytical estimation of possible strengthening showed that the yield strength of the nanobainite obtained should be in the gigapascal range, and the tensile testing results demonstrated that a 2 GPa yield strength was developed in the sample after isothermal treatment at 250 °C for 1 h. Investigations of the fracture surface and microstructure of the cross-section near the fracture zone confirmed the ductile mode of failure.

1. Introduction

High-strength materials are vitally important for human safety in different critical situations, including natural or man-made disasters. In this context, the use of materials with high strength significantly reduces the costs of producing load-bearing structures, increasing their reliability and structural strength. Numerous materials are characterized by high strength, including concrete [1], different kinds of composites [2,3], metallic glasses [4,5], and crystalline metallic alloys. The estimated strength can be on the order of tens of megapascals for concrete up to several gigapascals for composites, metallic glasses, and crystalline metallic alloys.

Besides the strength itself, the structural component should be able to withstand loading after the strength limit is achieved. This capacity is very important in cases in which the load increases suddenly for a short time. In these cases, some materials break in brittle mode while others just deform without breaking, with the latter having a certain reserve in load-bearing capacity that is important in diverse disaster scenarios.

Steel materials have such a reserve in load-bearing capacity. The strongest representative of this group is heavily drawn steel with eutectoid or near-eutectoid composition after isothermal treatment at a temperature near 500 °C. The ultimate tensile strength (UTS) of such material exceeds 6 GPa [6] due to lamellae structure refining [7], cementite partial decomposition [8], and its transition to an amorphous state [9]. However, such properties are achieved only for relatively thin sections (wires). Therefore, the practical applicability of this material is limited. The UTS of widely used steels for different constructions generally fluctuates in the range of 200–1700 MPa.

For a long time, significant efforts have been devoted to the investigation and development of high-strength steels. These steels are generally categorized according to their UTS. Steels with UTS up to 800 MPa are designated as high-strength steel (HSS) and advanced high-strength steel (AHSS), those with UTS over 800 MPa are designated as ultra-high-strength steel (UHSS) and, finally, steels with UTS over 1000 MPa are referred to as gigapascal steels [10]. High strength alone is insufficient and should be accompanied by acceptable ductility. Therefore, a widely accepted measure of structural strength of HSS, AHSS, UHSS, and gigapascal steels is the product of the ultimate tensile strength (UTS, GPa) and relative elongation (E, %) of a sample after testing (PSE, GPa × %).

Three generations of AHSS have been developed to date [10]. The first one includes dual phase (DP) steels, complex phase (CP) steels, high-strength low-alloyed (HSLA) steels, martensitic steels, and a low-alloyed steels with a transformation-induced plasticity (TRIP) effect [11,12]. Generally, first-generation AHSSs possess moderate mechanical properties (PSE up to 20 GPa × %). Second-generation AHSSs comprise high-alloyed austenitic single-phase steels, including twinning-induced plasticity (TWIP) steels, lightweight steels with induced plasticity (L-IP steels), and corrosion resistant steels with instable austenite structure [13], which demonstrate high PSE (up to 60 GPa × %) at UTS in the range of 600–1100 MPa. The practical use of second-generation AHSSs is limited due to their high alloying level and, hence, expensiveness. Third-generation AHSSs fill the gap between the first- and second-generation steels; i.e., their target PSE is about 30–40 GPa × %. Furthermore, their cost is reduced in comparison with the second-generation AHSSs.

The development of third-generation AHSSs is focused on achieving multiphase microstructures with elevated levels of retained austenite, which is prone to deformation-induced phase transformations. The transformation of metastable austenite into martensite under mechanical loading triggers the TRIP effect, significantly enhancing the strength-ductility balance. The TRIP effect is widely utilized in various steel grades to improve their mechanical properties and wear resistance [14,15,16,17,18,19,20,21,22,23]. Specialized heat treatments and thermomechanical processing can further optimize the performance of TRIP-assisted steels [24].

The general level of strength achieved by AHSSs is illustrated in Figure 1, which was compiled according to the data from works published in 2024–2025. Red dots correspond to maraging steels.

Of note, 2700 MPa UTS was achieved with high-carbon steel (1.0 wt.% C) after martensite quenching and tempering at 175 °C for 40+ days [25]. In addition, 60Si2CrVNb steel demonstrated 2578 MPa UTS after austempering [26]. Moreover, 2550 MPa UTS was noted for steel containing 0.53 wt.% C subjected to combination of isothermal treatment and plastic deformation [27]. Samples of modern UHSS 300M (0.42 wt.% C) obtained using laser powder bed fusion demonstrated 2500 MPa UTS after quenching and tempering [28]. In general, a UTS level of 2000–2500 MPa is the maximum achieved for steels with different levels of alloying, excluding eutectoid steels after patenting and plastic deformation, as previously mentioned.

Carbide-free bainitic steels are notable representatives of third-generation AHSSs [126]. The strengthening principle employed here involves treating the steel to achieving a nanostructured state with lower bainite during isothermal holding of supercooled austenite in the temperature range of 200–300 °C. An enhanced concentration of Si (above 1.5 wt.%) eliminates the precipitation of cementite during heat treatment [127]. As such, the resulting microstructure is a mix of bainitic ferrite and carbon-enriched austenite. Carbide-free lower bainite demonstrates high strength with acceptable elongation [128,129,130], but several difficulties have been noted regarding the wide implementation of austempering technology. Because the martensite-start temperature (Ms) is generally far above 20 °C, special baths are needed to eliminate or minimize martensite formation during the cooling of austenite in a single-phase region. The necessity of such baths significantly complicates the processing route, and the maximal dimensions of treated parts are limited. In addition, a long time is needed to obtain a decent quantity of lower bainite.

Spring steels alloyed by silicon are promising candidates for austempering due to silicon’s ability to suppress cementite formation. Numerous works, such as [26], achieve high strength after austempering of spring steels. Since higher carbon content is responsible for enhancing strength (Figure 1) and minimal alloying reduces treatment time, it is important to know the strength potential of simple high-carbon spring steel with minimal alloying. Therefore, this work deals with 60Si2 spring steel, which is analogous to 60si7 (DIN/EN) steel.

A temperature of 250 °C is important for isothermal holding because a lower temperature (200 °C) leads to a significant increase in the isothermal holding time and a decrease in the ductility of steel after heat treatment. In addition, a higher temperature (300 °C) leads to incomplete bainite transformation—i.e., higher amount of austenite—and therefore decrease in yield strength and UTS. Hence, austempering at 250 °C is expected to provide high YS and UTS with satisfactory elongation at a practically acceptable austempering time.

Thus, the goal of this work was to determine the optimal austempering time at 250 °C for unalloyed commercial spring steel with 0.6 wt.% C and enhanced Si content (60Si2 spring steel, analogous to 60si7 steel) and determine the microstructure and mechanical properties (YS, UTS, and elongation) for this steel after austempering at 250 °C.

2. Materials and Methods

A commercial rod of 60Si2 steel (Ø 16 mm) with the chemical composition shown

Table 1. Chemical composition of steel 60Si2, wt.%.

C	Si	Mn	Cr	S	P
0.60	1.73	0.72	0.03	0.002	0.016

was used for the experiments. The chemical composition of the steel was provided in a certificate from the supplier (ArcelorMittal, Kryvyi Rih, Ukraine).

Hardness tests were preliminarily performed to determine the necessary time for austempering. Seven different austempering treatments were performed with holding times of 4, 8, 15, 30, 45, and 60 min as described below. For each treatment, one control sample with a thickness of 5–6 mm was cut from steel rod. After the corresponding heat treatment, every sample was used in hardness tests performed with a Vickers hardness tester TVP-5012 (TOCHPRIBOR, Ivanovo, Russia) at a load of 98.1 N. A control hardness plate with a hardness of 816 HV10 was used to verify the hardness tester. Each hardness result was obtained as a mean of six to ten diagonal measurements. The 95% confidence limit for each hardness result did not exceed 5%. The hardness of a cross-section of a broken tensile sample was measured using a computer-controlled KB30S tester (Hegewald & Peschke, Nossen, Germany) under a load of 9.81 N. Tensile test samples, with a 5 mm diameter and a 25 mm gauge length, were machined from the steel rod. Tensile testing was conducted using a 100 kN electromechanical testing machine (UIT STM 100, Ukrintech Ltd., Kharkiv, Ukraine) under a speed of 2 mm/min without an extensometer. Before tensile tests, the testing machine’s dynamometer was verified against a control certified dynamometer. The error of load measurement was 0.5% or less. Diameters of tensile test samples were measured with accuracy of 0.05 mm. The samples underwent heat treatment. The treatment consisted of austenitization at 900 °C for 20 min, followed by austempering in a Pb-Sn bath at 250 °C for 1.0, 1.5, or 2.0 h and subsequent water quenching. Final cooling in water was used to avoid the appearance of possible unwanted structural constituents during slow cooling after austempering. Samples were prepared for microstructural analysis using standard techniques, involving grinding with abrasive papers of decreasing grit size, polishing with diamond paste, and etching in a nital solution. The microstructure and fracture surfaces were examined using a scanning electron microscope (SEM, JEOL JSM-7000F, JEOL Ltd., Tokyo, Japan) with acceleration voltage of 15 kV and magnifications up to ×30,000. SEM was performed because bainite can be clearly resolved under magnifications much higher than that achievable for optical microscopy. Microstructure images were analyzed with the ImageJ software. X-ray diffraction (XRD) analysis was performed using a DRON-3 diffractometer (Bourevestnik JSC, St. Petersburg, Russia) with Mo-Kα radiation.

3. Results

The initial microstructure of the as-received 60Si2 steel (Figure 2) comprised mostly an eutectoid constituent with a minor quantity of ferrite located on the boundaries of pearlite colonies. During heat treatment, the pearlitic phase completely disappeared at the initial stage (austenitization) and did not appear in subsequent stages.

The major parameters of austempering are the temperature of isothermal holding (Th) and holding time (HT). It is important to maintain Th within narrow limits; namely, above the martensite start temperature (Ms) for austenite, but not too high. It is recognized that increasing Th generally decreases the strength of austempered steel [131].

Different empirical models provide polynomial equations to determine the correlation between Ms and the chemical composition of austenite [132]. However, direct measurements of Ms provide more accurate values, which is important for the determination of Th. For example, dilatometer measurements for DIN1.5025 grade steel 51Si7 (0.53 wt.% C) provides Ms values of 273–275 °C [133,134]. Of note, 60Si2 steel differs from 51Si7 in carbon content, which is increased by 0.1 wt.%. According to empirical equations obtained from different sources [132], the carbon content variable differs, displaying a range of −302 to −584. Therefore, 0.1 wt.% C decreases the Ms temperature by 30–58 °C in comparison with that for 51Si7 steel. Hence, the Ms temperature for austenite of 60Si2 steel after homogenization in single-phase region is estimated to range from 217 to 245 °C. In [135], Ms = 250 °C was provided for a steel composition that is analogous to 60Si2 steel. Thus, the Th chosen for isothermal treatment of experimental samples was 250 °C.

The austempering time was determined based on the following considerations. A mixture of different phases in the microstructure can be obtained by cooling from the austenitization temperature to 250 °C, followed by isothermal holding for variable HT with subsequent final cooling in water. With increasing HT, the bainitizing process is initiated to a certain extent, and austenite is enriched with carbon [136]. This leads to a decrease in Ms, an increase in the content of austenite after treatment, and a decrease in final hardness. At some critical HT, the hardness should reach its minimum because of the two concurrent processes. The first process involves decreasing the amount of martensite, which leads to a decrease in hardness. The second one process involves increasing the amount of bainitic ferrite, which leads to an increase in hardness.

At some HT during isothermal treatment, the carbon content in austenite should increase to the amount that impedes martensite formation upon final cooling. In [132], a refined model for Ms calculation is developed for low Ms values:

Ms = 530.2 − 290.3 × [C] − 35.5 × [Mn] − 6.8 × [Si] − 20.8 × [Cr] − 17.2 × [Ni] − 10.4 × [Mo] + 7.1 × [Al] + 4.8 × [Co] − 75 × (1-exp(−0.96 × [C]))

(1)

Calculations according to Equation (1) [132] show that Ms decreases to 31 °C when the carbon content in austenite reaches 1.4 wt.%. A further increase in the carbon content to 1.5 wt.% causes Ms to decrease to 0 °C. Therefore, no martensite should appear after final cooling when carbon concentration in austenite is 1.4–1.5 wt.%. The subsequent increase in HT should lead to an increase in hardness due to the ongoing bainitic transformation and a decrease in the content of austenite. Finally, the increase in hardness should stop when the end of bainitic transformation is reached.

A similar experiment was conducted in [136]. DIN1.5025 grade 51Si7 steel (0.53 wt.% C) was cooled from the austenitization temperature to 350 °C and held for different HTs with final water cooling. The following HTs (in seconds) were used: 5, 30, 200, 600, 3600. Other samples were subjected to direct water quenching from the austenitization temperature. XRD investigations were performed to determine the austenite content and the carbon content in austenite after heat treatment. Figure 3 shows that, according to the data presented in [136], the austenite content in 51Si7 steel peaked after holding at 350 °C for 200 s. The carbon content in austenite reaches 1.36% after 600 s HT, and no significant increase was noted with further holding. Thus, in this experiment, the bainitic transformation ceased after 10 min of isothermal holding. The maximum content of austenite was attained before the transformation was stopped.

Austempering of 60Si2 steel at 250 °C should exhibit similar kinetic of changes in austenite and carbon contents. The significant differences included increased carbon content in steel and decreased temperature of isothermal holding. Since carbon diffusion is significantly slower at 250 °C than at 350 °C, it was expected that bainite transformation at 250 °C would be prolonged. According to [135], the bainite transformation in 60Si2 steel at 250 °C begins after 90 s (1.5 min) and ceases after approximately 10⁴ s (2.8 h).

Samples of 60Si2 steel were heat treated according the following protocol: austenization at 880–900 °C (20 min holding), isothermal holding at 250 °C with different HTs, and final cooling in water. The HTs were as follows (min): 4, 8, 15, 30, 45, 60. Direct water quenching from the austenitization temperature was also performed. Figure 4 demonstrates the influences of HT on the hardness of the heat-treated samples. According to Figure 4, the initial increase in HT leads to a significant reduction in hardness. Hardness reached a minimum after 15–30 min of holding, with further increases after 40–60 min of holding.

XRD data of thermally treated samples after direct quenching and isothermal holding (Figure 5) correlate with hardness measurements. After direct quenching, the sample shows no austenite peaks; this indicates the prevalence of martensite. The hardness of the samples after isothermal holding decreased significantly (Figure 4), which was attributed to the appearance of austenite in the microstructure (Figure 5). The calculated volume fractions of austenite for samples after austempering are shown in

Table 2. Effect of holding time at 250 °C on the volume fraction of retained austenite in 60Si2 steel.

Holding Time, min	Austenite, vol.%
0	0
4	9.5
8	10.9
15	11.8
30	10.1
45	8.8
60	8.8

. The maximum amount of austenite was reached after 15 min of holding. Therefore, based on the results obtained, it was established that the holding time of 60Si2 steel at 250 °C should be at least 1 h, consistent with data in the literature [136].

Of note, the diffractograms shown in Figure 5 only indicate the types of crystal lattices, i.e., α (BCC lattice-martensite or bainitic ferrite or both) and γ (FCC lattice-austenite). The corresponding diffraction maximums are marked as α or γ. Diffractogram 1 (Figure 5) corresponds to direct quenching; therefore, the maxima in the α diffraction correspond to martensite. After isothermal holding for 1.0 h and longer, the maxima in the α-diffraction correspond to bainitic ferrite.

SEM images of 60S2A steel at different magnifications after austempering at 250 °C for 1 h are shown in Figure 6. As expected, the carbide-free bainitic microstructure was developed. A higher magnification of ×30,000 revealed the nano-sized objects, which are bainite laths with an interlath distance within the nano-scale range.

Digital filtering of the image shown in Figure 6b reveals the interlath distance much more clearly (Figure 7a). Thus, it is possible to measure the interlath distance in nanobainite regions and the general distance between the phase borders with acceptable accuracy. The ImageJ software was employed to perform several hundreds of measurements of distance between the phase borders in semi-automatic mode. The dataset obtained was used to build the histogram and density plot of the distribution of inter-border distances (Figure 7b). The vast majority of distances between the phase borders occupy the nano-scale range with the mode in the range of 20–40 nm.

Figure 8 and

Table 3. Mechanical properties of 60Si2 steel after austempering for different time at 250 °C.

Holding Time, h	YS, MPa	UTS, MPa	Relative Elongation, %
1.0	2090 ± 94	2198 ± 99	9.0 ± 0.7
1.5	1886 ± 85	2163 ± 97	11.0 ± 0.9
2.0	1748 ± 79	2115 ± 95	16.9 ± 1.2

demonstrate the results of tensile tests with 60Si2 steel samples after austempering at 250 °C for 1, 1.5, and 2 h. The yield strength of the 1 h sample slightly exceeds 2000 MPa with 9% total elongation (Figure 9a). A gradual decrease in yield strength is established with increasing HT. This can be explained by thickening of bainite laths during excessive exposure at 250 °C [137]. The corresponding increase in interlath distance leads to a decrease in strength. The strength hardening rates (SHRs) obtained during the plastic deformation of the samples (Figure 8b) confirm the TRIP effect, especially in the sample after austempering for 1 h. The increasing austempering time leads to less pronouncing strengthening. Of note, after austempering for 1.5 and 2.0 h, the samples demonstrate positive SHRs with up to 10% or greater deformation. In certain cases, such increased plasticity and an extended TRIP effect can be more important that a slight decrease in YS. Thus, after austempering at 250 °C, 60Si2 steel possesses a very useful combination of YS, ductility, and positive SHRs over a broad range of austempering times.

Fracture surface of tensile sample (HT = 1 h) demonstrates the cup-cone pattern (Figure 9), which is typical for ductile steels [138,139]. In the central zone, the nucleation of micro voids proceeds with subsequent growth and coalescence [140]. The inclined surrounding fracture surface, referred to as shear lips [141,142], appears as a result of the combined effects of shearing and the normal detachment of a material [139]. The shear lips on the fracture surface (Figure 9c) exhibit dimples as a sign of pure ductile fracture. The fracture surface of the central zone (Figure 9d) still shows significant dimple areas (1), although multiple sites with quasi-cleavage patterns [143,144] are also present (2). The difference in the fracture surface of the central and peripheral zones indicates that these zones break in different stress states.

The longitudinal cross-section of the broken half of the sample (HT = 1 h) illustrates the ability of the bainite microstructure to undergo significant plastic deformation before fracture (Figure 10). Voids are visible in Figure 10a. Round fracture particles and particles with round edges are displayed in Figure 10b. This form of fracture particles correlates with general void-induced ductile fracture mechanism. The distribution of hardness across the broken tensile sample is shown in Figure 11.

4. Discussion

The question arises whether the level of YS obtained for 60S2A is consistent with theoretical concepts.

Numerous sources have suggested an analytical relationship between the yield strength of steels and the cumulative contribution of various parameters [145]. Applying this relationship to carbide-free nanobainitic microstructure yields the following equation:

YS = σ_o + ∆σ_SS + ∆σ_D + ∆σ_GB

(2)

where σ_o represents the Pierls–Nabarro stress, and ∆σ_SS is solid solution strengthening due to interstitial and substitutional atoms. In addition, ∆σ_GB, ∆σ_D, and ∆σ_P are grain boundary strengthening, dislocation strengthening, and precipitation strengthening, respectively. The remaining components of this relationship are pearlite strengthening, strengthening due to dispersed precipitates, and martensite/austenite islands. These factors are not considered here, as they are thought to be insignificant in carbide-free bainitic microstructures.

The Pierls–Nabarro stress value is 2 G × 10⁻⁴ MPa [146]. For iron, G = 84,000 Mpa. Therefore, the Pierls–Nabarro stress can be estimated as 17 MPa.

Solid solution strengthening is expressed as follows:

$$∆{\sigma }_{SS}=\sum _{i=0}^n{k}_i{c}_i$$

(3)

where k_i is the coefficient for element i in solid solution and $c_i$ represents the concentration for that element (in wt. %). If we suppose a uniform distribution of silicon and manganese between bainitic ferrite and austenite [147], then solid solution strengthening for bainitic ferrite can be estimated based on silicon (1.73 wt.%) and manganese (0.72 wt.%) concentrations. According to [147], the coefficients for C, Si, and Mn strengthening are 5440, 83, and 32 MPa·wt%⁻¹, respectively. Thus, solid solution strengthening is assessed as 144 MPa for Si and 23 MPa for Mn. In [137], the carbon content in bainitic ferrite after long austempering of high-carbon steel at 250 °C was estimated as 0.12–0.21 wt.%. Given a carbon content of 0.16%, the carbon contribution to strengthening of bainite is approximately 900 MPa.

Dislocation strengthening is generally calculated as follows [148]:

∆σ_D = αMGbρ^1/2

(4)

where α = 0.5, M = 3.06 (Taylor factor), G = 84.000 MPa (shear modulus), b = 2.5 × 10⁻¹⁰ m for iron (Burgers vector), and ρ denotes dislocation density. In [149], the dislocation density in bainitic ferrite was estimated as 2.1 × 10¹³ m⁻² to 7.1 × 10¹³ m⁻². Thus, dislocation strengthening contributes on average 217 MPa to the yield strength value.

The total effect of the above three factors on the level of yield strength yields 1134 MPa.

Strengthening from microstructure refinement can be assessed using the Hall–Petch equation, as follows:

∆σ_GB = k_yd^−1/2

(5)

where k_y = 0.63 MPa·m^−1/2 [146] and d is the average grain size. The strengthening from microstructure refinement is associated with hinderance for dislocation sliding with decreasing grain size; i.e., in our case, the interlath distance.

A graphic interpretation of Equation (5) shows (Figure 12) that the strongest strengthening effect comes from microstructure refinement to the nanoscale interlath distance.

According to the interlath distance distribution (Figure 7), the mode distance is 30 nm or 0.03 µm. The Hall–Petch equation returns ∆σ_GB value in gigapascals range for this distance. Of note, different references provide different values of k_y coefficient; however, the resulting ∆σ_GB is not less than 1000 MPa.

Thus, the theoretical estimation of the yield strength for 60Si2 steel after austempering is 2000 MPa.

The results presented clearly illustrate the possibility of obtaining the nanobainite microstructure in standard high-carbon spring steel with minimum alloying. In fact, only enhanced carbon content and alloying with silicon are absolutely necessary for carbide-free nanobainite development. Other alloying elements significantly reduce the bainitic transformation. Additional alloying is required to fine-tune Ms and reduce the diffusion transformation of austenite in the temperature range of 500–600 °C. This information should improve the convenience of the technological processing of actual industrial components.

5. Conclusions

A carbide-free nanobainite microstructure was obtained in low-alloyed spring steel 60Si2 (60si7) via austempering at 250 °C for up to 2 h. The tensile test results demonstrated the significant potential of spring steel after austempering as a high-strength material. The conclusions drawn from presented work are as follows:

• The minimal content of alloying elements provides the possibility to obtain a nanobainite microstructure in 60Si2 steel at a low temperature of austempering during a relatively short time. This time is comparable to the generally accepted time for heat treatment of steel parts, such as that for tempering. XRD investigations demonstrated that the incubation period for bainitic transformation in 60Si2 steel at 250 °C is less than 4 min.
• The distribution of distances between phase borders in bainite indicate that the mode value is 30 nm. According to the Hall–Petch law, such a nanoscale interlath distance provides yield strength at the gigapascal level.
• Along with the high strength obtained, an acceptable level of ductility was also achieved. Analysis of the fracture surface and microstructures of the longitudinal cross-section of the broken sample confirmed the ductile mode of fracture.
• Further research should focus on optimizing the composition and processing routes of spring steels for use as high-strength materials in safety-critical structural applications.