Introduction

Interstitial-free (IF) steel has found broad applications in the automotive sector for fabricating body panels. The high formability and nonaging behavior make it advantageous due to its easy drawability into complex shapes.1 To reduce the carbon footprint of vehicles, a material with high strength-to-weight ratio for better fuel efficiency and to fulfil safety norms is desired. Many of the commercially available grades of steel [i.e., dual phase and high-strength low-allow (HSLA)] used in the automotive sector show high strength but lack formability. The strength of IF steel may be enhanced by either alloying or mechanical working. Precipitation hardening with Cu,2,3,4 nitriding with potassium nitrate salt,5 and strengthening with P6,7 and B8 are efficient methods for enhancing the strength of IF steel. Apart from precipitation strengthening of IF steels, attempts have been made to strengthen this steel by cold working and applying various severe plastic deformation (SPD) techniques such as accumulative roll bonding (ARB), equal-channel angular pressing (ECAP), surface mechanical grinding treatment (SMGT), high-pressure torsion (HPT), laser shock peening (LSP), surface mechanical attrition treatment (SMAT), etc.8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26 IF steel cold rolled to 90%, 95%, 98%, and 98.5% reduction in thickness showed peak hardness of 1.46 GPa, 1.62 GPa, 1.85 GPa, and 3.13 GPa, respectively.8 Several investigations have been carried out to enhance the yield strength (YS) and ultimate tensile strength (UTS) of IF steel through accumulative roll bonding (ARB). However, in all those investigations, the total elongation (ɛt) was found to be very poor.15,16,17 The strengthening was attributed to grain boundary and dislocation strengthening, but these ARB-processed materials exhibit a strength–ductility trade-off (ɛt  < 2%). Several investigators have reported the effect of ECAP on the strength and ductility of IF steel.9,11,27,28,29,30 Although the increase in the YS of IF steel after ECAP was almost 4–6 times higher than that of coarse-grained (CG) materials, the ductility was reduced and in most cases low (ɛt  < 0%). Tao et al. reported that it was difficult to synthesize bulk nanostructured materials without porosity and contamination, and also to establish a structure–property correlation without any external influencing factor.31

It is well known that the failure of materials under fatigue, fretting fatigue, corrosion, and wear takes place from the surface.32 It was found that surface nanostructured materials are superior to conventional coarse-grained polycrystalline materials.33 Nanocrystalline materials are superior in terms of hardness and strength,26,31,32,34,35 enhanced physical properties,36,37,38 improved tribological properties,39 etc. The SPD process for surface nanostructuring is now a well-established method for microstructural modification, enhancing mechanical properties and improving the functional properties of materials. In our previous investigation, LSP was applied on the IF steel surface to generate gradient-structured IF steel with an ~ 84% increase in YS and substantial elongation (ɛt > 35%).25 IF steel processed through surface mechanical grinding treatment (SMGT) formed a gradient nanostructured layer with grain size up to 20 nm and peak hardness of 5.3 GPa on the treated surface.34 Similarly, SMATed IF steel with thickness of 1 mm showed 2.6 times higher YS than as-received IF steel.26 Although a few literature sources on surface nanostructuring of IF steel are available, there are no systematic studies reporting the effect of SMAT on the surface roughness, microstructural modification, microhardness gradient, tensile properties, and thermal stability of IF steel.

In the present investigation, efforts were made to investigate the effect of the duration of SMAT on the surface roughness, microstructural modifications, phase composition, microhardness gradient, tensile properties, and thermal stability of IF steel.

Experimental Procedures

Material and SMAT

Interstitial-free steel sheet with dimensions of 200 mm × 200 mm × 4 mm was procured from M/s Tata Steel, Jamshedpur, India. The IF steel used in the present investigation contains 0.004% C, 0.01% Si, 0.0623% Mn, 0.153% Al, 0.042% Cu, 0.006% Nb, 0.068% Ti, 0.0038% N, and balance Fe (wt.%). The details of sample preparation for SMAT are given elsewhere.25 An ultrasonic shot peening (USSP) machine (SONATS STRESSVOYAGER®, France) was used to SMAT the IF steel using different processing parameters (Table I). The samples SMATed for 200 s were further stress relieved at 300°C and 500°C for 60 min as per the procedure mentioned elsewhere.40

Table I Processing parameters of SMAT
Full size table

Material Characterization and Mechanical Properties Evaluation

To examine the surface profiles of the specimens before and after SMAT, a surface profilometer (SURFTEST - SJ 410; Mitutoyo, Japan) was used. Visible-light microscopy (VLM, Imager.M1m AX 10 model; Zeiss) was used to characterize transverse sections of the as-received/unSMATed and SMATed samples. SEM (QUANTA 200F, 30 kV) was used to study the surface morphology and microstructural refinement following SMAT. To examine the fine microstructural features of the SMATed surfaces, TEM (TECNAI G220, 200 kV) was used. Electron-transparent samples were prepared by dimple grinding (GATAN Dimple grinder-II) and ion milling (GATAN 695, PIPS-II).

The hardness was determined using both macro- (Vickers hardness testing machine, Leco LV700AT; load 98.07 N; dwell time 10 s) and microhardness (Shimadzu; load 430.05 mN; dwell time: 5 s) testing, using at least ten readings (in depth from the SMATed surface) to calculate the standard deviation. The details of the tensile testing and sample specification are described elsewhere.25

Results

Surface Topography and Morphology

The average surface roughness (Ra) was evaluated by measuring the vertical deviations of the roughness profile from the mean line using a Mitutoyo surface profilometer. UnSMATed IF steel showed a lower Ra value compared with the SMATed samples (Table II). The effect of SMAT on the treated specimens was observed in terms of craters and humps in Fig. 1b, c, and d and increased Ra (Table II). Similar results were found by Pandey et al. for Al alloys.36,41 The Ra value increased with increasing duration of SMAT from 100 s to 300 s. However, it is evident that the Ra value for durations of 100 s and 200 s were similar, while that for 300 s was significantly higher. The increase in the Ra is due to the increased duration of SMAT.

Table II Variation of average surface roughness in different treated condition(s)
Full size table
Fig. 1
figure 1

(a) Visible-light micrograph of unSMATed IF steel, and SEM micrographs of IF steel samples SMATed for (b) 100 s, (c) 200 s, and (d) 300 s showing surface topography.

Full size image

Microstructural Characterization

The microstructure in the unSMATed, SMATed, and SMATed + stress relieved (SR) conditions was examined by VLM, SEM, and TEM. The microstructure of the untreated sample showed equiaxed grains of ~ 29 µm mean intercept length (Fig. 1a). The microstructure of the SMATed surfaces without and with stress relieving treatment is shown in Fig. 2a, b, c and d but is very difficult to resolve by VLM. The surface modification from the top of the treated surface towards the interior is revealed by the optical micrographs of the SMATed IF steel. SEM micrographs at higher magnification at a certain depth from the top of the treated surface towards the interior confirmed the presence of nanograins in the transverse section of the specimen in Fig. 2e, f and g. Figure 2b shows a typical gradient microstructure consisting of nanograins (NG), ultrafine grains (UFG), and coarse grains (CG) from the surface towards the interior. The surface modification was found to be dependent on the process variables of SMAT as listed in Table I. Increasing the duration of the treatment increased the depth of the surface modification. The depth of surface modification for the IF steel SMATed for 100 s, 200 s, and 300 s was found to be ~ 230 µm, ~ 463 µm, and ~ 510 µm, respectively. The depth of surface modification even after stress relief at 500°C for the sample SMATed for 200 s was found to be ~ 180 µm. This appreciable depth of refinement for the stress-relieved samples establishes the thermal stability of the nanostructured layer. An NG regime with thickness of 24 µm, 88 µm, and 40 µm is evidence for the samples SMATed for 100 s, 200 s, and 300 s, respectively, in Fig. 2e, f and g. Prominent cracking was observed in the sample treated for 300 s, as discerned from Fig. 2g.

Fig. 2
figure 2

Cross-sectional visible-light micrographs from the top of samples SMATed for (a) 100 s, (b) 200 s, (c) and 300 s, and (d) SMAT for 200 s and stress relief at 500°C; Cross-sectional SEM images from the top of the samples SMATed for (e) 100 s, (f) 200 s, and (g) 300 s.

Full size image

The evolution of the microstructure during SMAT of the IF steel can be better understood by TEM. The TEM micrograph in Fig. 3 shows the fine microstructural features of the samples treated for 100 s, 200 s, and 300 s. The selected-area diffraction patterns (SADPs) show polycrystalline rings due to the formation of equiaxed NG during SMAT. The continuity of the ring increases as a function of the duration of SMAT due to an increase in random crystallographic orientation. The rings in the SADP correspond to the (110), (200), (211), (220), and (310) crystallographic planes of α-Fe. Continuous rings were evident in the SADP of the IF steel SMATed for 100 s (Fig. 3a), and the corresponding bright-field image confirms the presence of dense dislocation walls (DDWs) and dislocation tangles (DTs) after SMAT. Figure 3d shows a dark-field image corresponding to spot 1 marked in the SADP in Fig. 3c. The presence of nanostructured grains on the treated surface is evident in Fig. 3d. However, the samples SMATed for 200 s showed the formation of subgrain boundaries due to the random impingement of balls during processing, as seen in Fig. 3f and g, which further led to grain refinement. Furthermore, increasing the SMAT duration to 300 s led to the formation of fine nanostructured grains with high crystallographic misorientation (Fig. 3i, j, k and l). Refinement in the nanostructured grains is evident with increasing duration of SMAT. As listed in Table III and shown in Fig. 3d, g, h, k and l, the NG size was ~ 27 nm, ~ 21 nm, and ~ 18 nm for the IF steel SMATed for 100 s, 200 s, and 300 s, respectively.

Fig. 3
figure 3

(a) SADP of IF steel SMATed for 100 s; (b) bright-field image corresponding to SADP; (c) SADP and bright-field image from different area; (d) dark-field image corresponding to spot 1; (e) SADP of IF steel SMATed for 200 s; (f) bright-field image corresponding to SADP; (g) SADP and bright-field image from different area; (h) dark-field image corresponding to spot 1; (i) SADP of IF steel SMATed for 300 s; (j) bright-field image corresponding to SADP; (k) dark-field image corresponding to spot 1; (l) dark-field image corresponding to spot 2.

Full size image
Table III Crystallite size, lattice strain, lattice parameter, dislocation density, and grain size of SMATed IF steel
Full size table

Phase Analysis

Figure 4 shows the x-ray diffraction (XRD) profiles of the unSMATed and SMATed IF steel. Only peaks for α-Fe are observed, with no observable phase transformation after SMAT of the IF steel. However, the Bragg diffraction peaks after SMAT became broader than those of the sample before SMAT, and a peak shift was evident, as discerned in Fig. 4b showing the (110) reflection of SMATed IF steel. The broadening of the Bragg peaks indicates grain refinement and an increase in the lattice strain with increasing duration of SMAT. The crystallite size of the SMATed samples was calculated from the full-width at half-maximum (FWHM) of four Bragg reflections corresponding to α-Fe: (110), (200), (211), and (220), using the Williamson–Hall equation.42

Fig. 4
figure 4

Phase analysis of unSMATed and SMATed IF steel (a); broadening of (110) peak for SMATed samples (b); effect of SMAT duration on dislocation density and lattice parameter of SMATed IF steel (c).

Full size image

The average crystallite size and the lattice strain of the treated samples were found to be approximately 44–23 nm and 0.229–0.259, respectively (Table III). The lattice parameter (a) and dislocation density (ρ) are also presented in Table III. With increasing SMAT duration, the lattice parameter ‘a’ decreased up to 300 s of SMAT and the dislocation density increases as observed in Fig. 4c and Table III. The dislocation density was calculated using the expression given below43:

$$ \rho = \frac{{3\sqrt {2\pi} < \varepsilon^{2} >^{1/2} }}{tb}, $$
(1)

where t and ε are the crystallite size and lattice strain reported in Table III, b is the Burgers vector of body-centered cubic (BCC) metal \( \left[ {b = \frac{a\sqrt 3 }{2}} \right] \).

It can be seen that the grain size determined by TEM was lower than that derived by XRD analysis (Table III). The TEM information was obtained from the 20-µm-thick SMATed layer, while the XRD information was averaged. This discrepancy between the grain sizes of the treated specimens can be attributed to the bulk and localized interaction of electromagnetic radiation in XRD and TEM analysis, respectively.

Hardness and Microhardness

The hardness of the SMATed surface was found to be 42–70% higher than that of the unSMATed samples. The bulk hardness of the SMATed sample was much higher than that of the unSMATed sample and increased with the duration of SMAT (Table IV). Figure 5a presents the microhardness profiles of the unSMATed, SMATed, and SMATed+SR (at 500°C) samples processed for different durations (Table I). It can be seen that, with increasing SMAT duration, the surface hardness increased and was highest for the 200-s treatment (Table IV). In almost all cases, the microhardness gradient was observed up to a depth of ~ 700 μm. The microhardness modification of the treated surface was observed up to a depth of ~ 915 μm for the sample SMATed for 200 s (Fig. 5a). The microcracks on the surface of the sample treated for 300 s might be responsible for the slight decrease in the peak surface hardness and depth of microhardness gradient (Table IV). Although the thickness of the microhardness gradient and peak surface hardness decreased after SR of the SMATed sample, it remained appreciably higher than that of the untreated samples. The gradient in the microhardness shown in Fig. 5a is quite consistent with the gradient microstructure shown in Fig. 2 having NG–UFG–CG regions.

Table IV Tensile properties of unSMATed and SMATed IF steel
Full size table
Fig. 5
figure 5

Microhardness gradient from top of treated surface for unSMATed and SMATed IF steel (a); multivariate radar diagram indicating mechanical properties of SMATed IF steel (b).

Full size image

Tensile Properties

It can be seen that the 0.2% offset YS increased by approximately 66–125% for the SMATed IF steel. As seen in Fig. 5b and Table IV, the YS and UTS increased up to 200 s of SMAT. However, there was no apparent increase in the YS or UTS thereafter, and they decreased for the 300 s SMAT and 200 s SMAT+SR samples. The multivariate radar diagram in Fig. 5b illustrates that the mechanical properties are dependent on the SMAT duration, with the most favorable properties being found for the 200 s SMATed sample. The YS and UTS increased significantly without much loss of ductility (Fig. 5b and Table IV). The percentage ɛt of the 200 s SMATed sample was almost comparable to that of the untreated sample. The YS and UTS of the 300 s SMATed samples were slightly inferior to that of the sample treated for 200 s. This can be attributed to the formation of surface cracks during SMAT (Fig. 2g), which decrease its load-bearing capacity. The mixed mode of failure is evident from the fractographs of the SMATed samples (Fig. 6a, b, c, d, e, f, g, h and i). Ductile fracture was evident only for the unSMATed IF steel (Fig. 6a and b), but the SMATed samples showed a mixed mode of failure, i.e., quasicleavage and ductile, which is evident from the gradient nanostructured layer seen in Fig. 6c, d, e, f, g, h and i. For the samples SMATed for 200 s and 300 s, quasicleavage failure with ductile failure are quite evident in Fig. 6. In the fractographs of the samples SMATed for 200 s, a few regions of shallow dimples are also evident, but the volume fraction of shallow dimples increased with the duration of SMAT for the 300 s SMATed IF steel. This transition in the failure mode from ductile to mixed mode suggests a composite microstructure of the SMATed IF steel.

Fig. 6
figure 6

Fractograph of IF steel (a) unSMATed at low magnification; (b) unSMATed showing dimples; (c) SMATed for 100 s; (d) SMATed for 200 s at low magnification; (e) SMATed for 200 s showing dimples; (f) SMATed for 200 s showing dimples in the coarse-grained region; (g) SMATed for 300 s at low magnification; (h) SMATed for 300 s showing quasicleavage; (i) SMATed for 300 s showing shallow dimples in the coarse-grained region.

Full size image

Discussion

The present study addresses the effects of SMAT on the surface roughness, microstructural refinement, hardness, and tensile properties of IF steel. In the case of SMAT, impingement of hard steel balls occurs in multiple directions, and the initial ridges formed are repetitively impinged upon by balls in the course of the treatment. The repetitive multidirectional SMAT treatment leads to an increase in the stress at high strain rates with an increase in the dislocation density. Due to the effective multidirectional impingement by hard steel balls, the coarse surface grains undergo fragmentation at the surface and become nanostructured due to SMAT. The average grain size of the nanograins and the thickness of the nanostructured surface layer depend on the SMAT parameters such as the ball size, amplitude of ultrasonic vibrations, and duration of treatment.32

SMAT modifies the surface up to a depth of ~ 510 µm, and a gradient NG–UFG–CG microstructure with nanostructured grains of ~ 18 nm develops at the top of the treated surface. Wu et al. and Liu et al. also reported a similar gradient microstructure of NG–UFG–CG in IF steel following SMAT26 and SMGT,34 respectively. They reported microstructural refinement up to a depth of 125 µm and 150 µm from the top of the treated surface by SMAT and SMGT, respectively. A similar kind of gradient microstructure was found in pure Cu23 and Ni.44 In many investigations, grain sizes of ~ 20–200 nm have been reported for IF steel subjected to cold rolling,4 ARB,15 and ECAP.29 Wu et al. reported NG of 96 nm size, without much mention in the variation of grain size with the duration of SMAT. In the present investigation, it is seen that the NG size varied from ~ 18 nm to 27 nm (Table III). In another study, Liu et al. reported an NG of 20 nm on a surface processed by SMGT.45 The variation in the duration of SMAT affected the thickness of the gradient microstructure, revealing better results than in previous literature reports on surface nanostructured IF steels.

To accommodate the plastic strain generated in IF steel during SMAT, different dislocation activities such as gliding, accumulation, and tangling normally occur.46 The mechanism of grain refinement and dislocation arrangement in metals and alloys depends on the crystal structure and stacking fault energy (SFE).46,47 The grain refinement and microstructural evolution in pure Fe with BCC crystal structure were well established by Tao et al.,47 who proposed a three-step mechanism for grain refinement of pure Fe: (1) development of DDWs and DTs for accommodation of plastic strain, (2) transformation of DDWs and DTs to subgrain boundaries with small misorientation, and (3) transformation of subgrain boundaries to highly misoriented grain boundaries. IF steel has BCC crystal structure, and the TEM micrographs in Fig. 3 are in agreement with the mechanism proposed by Tao et al.47 A change in the slip system with strain within the same grain during multidirectional peening of Fe was observed by Tao et al.46 During SMAT, dislocation interactions take place on the most favorable and other less active slip planes. The interactions of dislocations on active and less active slip planes may be held responsible for the more efficient subdivision of grains by DDWs and DTs in IF steel during SMAT as compared with other grain refinement processes such as ECAP, ARB, and HPT.

There is an appreciable enhancement in the hardness and microhardness of the IF steel subjected to different durations of SMAT (Fig. 5b, Table IV). Wu et al. reported that the peak microhardness was ~ 2.60 GPa and the gradient was observed only up to a depth of 100 µm.26 The peak surface microhardness is in accordance with the present investigation, but the thickness of the microhardness gradient from the top of the treated surface is clearly less compared with the present study. Even after SR treatment at 500°C of the IF steel SMATed for 200 s, the depth of the microhardness gradient was ≥ 700 µm. Zhang et al. reported the microhardness for the SMATed Fe to be ~ 1.70 GPa, and with an increase in the temperature of SR treatment to 600°C, the peak hardness of the Fe sample alloyed with Ni decreased to 1.31 GPa.48 However, in the present investigation, the peak microhardness even after stress relief remained comparable to those of SMATed samples.

The strengthening mechanism in BCC Fe proposed by Kamikawa et al. suggests that crystal defects act as obstacles to dislocation movement.49 In the present case, both the grain refinement and dislocation density are appreciable. Therefore, the grain refinement during SMAT and the increased dislocation density may be responsible for increasing the strength of the IF steel.

The strength increases up to 200 s of SMAT and is comparable to that of the sample SMATed for 300 s. The saturation in the YS and UTS for the sample SMATed for 300 s can be attributed to crack formation during longer-duration treatment. In almost all the investigations on SPD other than SMAT, it has been reported that the strength increased by approximately six- to eightfold, but the percentage ɛt was less than ~ 16%. However, in the case of SMAT, it is seen that the strength increases by approximately two- to threefold with appreciable ductility comparable to that of the unSMATed sample. Chen et al. reported an approximately sixfold increase in strength for 316L SMATed steel with an ɛt of approximately 3.4% for the sample with thickness of 1 mm due to bulk nanostructuring.50 However, in the present investigation, synergy is maintained between strength and ductility. Such synergy between strength and ductility for the nanostructured gradient layer was proposed by Ma and Zhu.33 They stated that the plastic strain gradient induces multiaxial stress states under uniaxial loading conditions. This arises due to the interplay between the fine- and coarse-grained layer due to strain incompatibilities. This increase in the YS with appreciable ductility is due to transient hardening in the regime of small tensile strain for the fine- and coarse-grained regions in IF steel.

Conclusion

The influence of SMAT on the microstructural refinement, gradient microhardness, and strength of IF steel was studied systematically. The following conclusions can be drawn from the present work:

  • IF steel exhibited microstructural refinement, gradient microhardness, and strength with significant ductility. There was an appreciable enhancement in the hardness and peak surface microhardness by ~ 70% and 43%, respectively.

  • The microhardness increased with increasing duration of SMAT. The SMAT effectively refined the microstructure up to a depth of ~ 900 µm (for the sample SMATed for 200 s).

  • There was a significant increase in the strength of ~ 125% (2.3 times), with ductility almost equal to that of the unSMATed specimen.

  • The nanostructured gradient layer of IF steel SMATed for 200 s was found to be thermally stable even at 500°C up to a depth of ~ 700 µm. The increase in the peak surface hardness of the SMATed IF steel after stress relief at 300°C and 500°C was found to be ~ 29% and 52%, respectively.