Evaluation of Strengthening Mechanisms in Novel Fully Ferritic Advanced High‐Strength Steels

New steel alloying concepts are designed in order to produce a fully ferritic, low‐alloy steel with high (1 GPa) ultimate tensile strength (TS). A simulated hot‐deformation process of the Ti–Mo–V–Nb and Ti–Mo–V steels is designed for that purpose, and the strengthening mechanisms of the steels are evaluated after the isothermal dwell at three different temperatures (590, 630, and 680 °C). The TS and the yield strength (YS) of the test alloys are estimated via hardness measurements. Results show that the estimated TS of over 1000 MPa and YS of over 900 MPa can be achieved in both steels, although the contribution of different strengthening mechanisms to the YS varies between the steels. The effect of the dislocation strengthening can especially compensate the reduced effect of the precipitation strengthening at all tested coiling temperatures (CTs). Based on the results, a CT range of 590–630 °C with the 1800 s dwell time seems to be a potential process window for the studied steels after the present thermomechanically controlled processing (TMCP) route.


Introduction
The current global need for more energy-efficient transportation solutions has generated extensive research work in the academic community and the steel industry.It has aimed at developing novel, advanced high-strength steels (AHSS) with high strength and excellent formability properties.However, the premature cracking during stretch flanging processes of AHSS has hindered their implementation in the automotive industry.Therefore, the effects of the thermomechanical-controlled rolling parameters on the microstructure, mechanical properties, and stretch flangeability of various low-alloyed steels have been widely studied during the last years.3][4] The automotive steel grades with the tensile strength (TS) less than 700 MPa typically have hole expansion ratio (HER) values over 60%. [5]The HER value depends mainly on the microstructure constituents and their mechanical properties, distribution, and volume fractions within the steel matrix when the TS of AHSS grades is higher than the just-mentioned value.8][9][10][11][12] Excellent strength and flangeability properties can be achieved in the complex-phase (CP) steels and in a steel consisting of tempered martensite and granular bainite; however, that required a strict thermomechanical process control together with a well-balanced steel chemistry. [2,13]esides the abovementioned steels, the fully ferritic steel grades typically have excellent formability but a relatively low strength; thus, Funakawa et al. [14] introduced a method to increase their strength substantially by nanometer-sized interphase precipitates (IP) that were formed during the austenite-to-ferrite phase transformation.They developed a new 0.04%C-1.5%Mn-0.09%Ti-0.2%Mosteel in which the fine (∅ = 3 nm) interphase carbides in rows existed within the fully ferritic matrix.A new 980 MPa fully ferritic, hot-rolled steel was later studied in the laboratory scale by Huang et al. [15] They also studied the effect of the coiling temperature (CT) on the precipitation behavior of the low carbon (C = 0.09 wt%), low-alloyed Ti-Mo steel using three different isothermal CTs and 1 h dwell time (DT).It was shown that the lowest CT of 600 °C was required for the excellent strength.
[18][19] Hutten et al. [16] investigated the mechanical properties of three low-carbon Ti-Mo-Nb-(V) steels.The coiling time required to produce a ferritic microstructure with less than 10 vol% martensite would take significantly more time for the Nb-Mo steel when compared with the Ti-Mo steels. [17]inally, a TMCP route of the low-alloyed MoCrNbV and VCrNb steels with low carbon content was simulated and its effects on the microstructure and nanoscale precipitates were investigated recently. [18,19]Microtensile tests revealed that over 800 MPa TS can be achieved in these predominantly ferritic steel alloys with some hard microstructural constituents (bainite and pearlite).
Considering the strengthening mechanisms and their contributions to the yield strength (YS) of the low carbon low-alloy steels, Kim et al. [20,21] showed that the maximum effect of precipitation hardening and grain-refinement hardening could not be reached simultaneously in a Ti-Mo steel.Thus, the optimal combined hardening effects by relatively fine ferrite grain size and nanosized interphase precipitate distribution could be only achieved by careful planning of TMCP procedures.
The TMCP procedures of these low-alloyed ferritic steels generally begin with the homogenization treatment at a temperature range of 1200-1250 °C in order to resolve the carbides.[26] The microstructure of the studied steels contained at least 10 vol% of hard microstructural constituents (bainite and pearlite), despite the promising precipitate hardening properties achieved using the investigated TMCP routes.
The contribution of different strengthening mechanisms to the YS of the low-alloyed Ti-Mo has been evaluated in earlier studies. [15,19,21]Together with the lattice friction stress of pure iron, the solid solution hardening, the grain refinement hardening, the precipitation hardening, and the dislocation hardening have been taken in consideration.The notable effect of the grain size and the precipitates on the YS has been verified in these studies, but the results concerning the efficiency of the lastmentioned mechanisms are controversial.Huang et al. [15] and Singh et al. [19] predicted that the value of the dislocation hardening was almost 170 MPa in their steels, whereas Kim et al. [21] stated that its contribution to the YS of the studied steels was negligible.This discrepancy needs clarification because the dislocation hardening may be one of the key elements if the strength properties of the fully ferritic steel are aimed to be equal with the 980 MPa steel grade.
Therefore, this study investigated the decomposition of austenite during the simulated TMCP route and the different coiling procedures in two fully ferritic low-carbon, low-alloyed steels.The effect of the coiling parameters and steel chemistry especially on the dislocation hardening and other strengthening mechanisms was thoroughly investigated.The TS and the YS of the test alloys were estimated via hardness measurements.The contribution of the main factors (grain size, solid solution, precipitation, and dislocation hardening) on the TS of the test alloys was evaluated using X-ray diffraction (XRD), field emission scanning electron microscope (FE-SEM)/electron backscatter diffraction (EBSD), and High-resolution transmission electron microscopy (HR-TEM) analyses.Based on these results, an optimal process window was suggested for the manufacturing of an ultrahigh strength steel sheets (TS ≥ 1000 MPa) with a novel steel chemistry.

Experimental Section
Table 1 gives the chemical compositions of the tested steel alloys in weight percent (wt%).The investigated steels were named Ti-Mo-V steel and Ti-Mo-V-Nb steel.The additions of microalloying have been calculated so that there should not be any excess carbon, for example, cementite formation.A lesser amount of Ti in material two has been compensated with a higher amount of V and the addition of Nb.Steels were produced as laboratory castings and were cast using a vacuum furnace.The dimensions of each laboratory cast were 125 Â 170 Â 420 mm (h Â w Â l) and weighed approximately 75 kg.The casts were cut into smaller pieces (50 Â 80 Â 150 mm) and hot rolled to a thickness of 10 mm using a soaking treatment of 1200 °C for 2 h to achieve material for Gleeble tests.
The simulated TMCPs and the dilatometric measurements were conducted with a Gleeble 3800 simulator using cylindrical samples with a diameter of 6 mm and a length of 9 mm.All samples were austenitized at a temperature of 1260 °C for 5 min at the first stage to resolve carbides.The specimens were slowly cooled in the next stage under their predicted nonrecrystallization temperature (T nr ) of 920 °C.The specimens were deformed at this temperature to a total strain (ε tot ) of 0.6 followed by fast accelerated cooling (50 °C s À1 ) to the CTs of 590, 630, and 680 °C.The DTs of 300 and 1800 s were used in each CT before the specimens were air cooled to the room temperature (RT).The specimens were coded as 590-300, 590-1800, 630-300, 630-1800, 680-300, and 680-1800 based on the CTs and DTs.The Ultimate tensile strength (UTS) evaluation of the test specimens measured their hardness using the Vickers method (load of 10 kg) with five measurements of each sample.The exact details of the hot deformation process are confidential.
Microstructures of the etched (Nital) TMCP specimens were characterized with a ZEISS Ultra Plus FE-SEM.EBSD was performed using AztecHKL acquisition and analysis software.The EBSD measurements were made using an accelerating voltage of 20 kV and a step size of 0.1 μm.The EBSD analyses were performed using AztecHKL acquisition and analysis software.A Rigaku SmartLab X-Ray diffractometer with a 5.4 kW Co radiation source and a CBO-f polycapillary X-Ray optics was used in the XRD data acquisition.The step size of 0.01°and the scan speed 0.5°/60 s over the 2-theta range of 45°-130°were used in the measurements to estimate dislocation densities and detection of a possible volume fraction of retained austenite in the specimens.The fundamental parameter (FP) method of the PDXL2 analysis software was used to estimate the specimens' crystallite sizes, microstrains, and lattice parameters.
The thin foil specimens for the TEM study were prepared using FEI Helios DualBeam FIBþFE-SEM/Scanning transmission electron microscopy (STEM) equipment.The morphology, chemistry, size distribution, and structural characterization of the precipitates were determined in the STEM/energy-dispersive spectroscopy (EDS) analyses.A high-resolution microscope JEOL JEM-2200FS Energy-filtered transmission electron microscopy (EF-TEM)/STEM with EDS microanalyzer were used with the operating voltage of 200 kV.The detailed description of the precipitation behavior of the test steels was reported earlier in ref. [27].

Microstructure and Grain Size
The dilatometric measurements confirmed that the austenite-toferrite phase transformation rate of the test steels was significantly higher at the 630 and 680 °C temperatures compared to the lowest temperature of 590 °C (Figure 1).Moreover, the austenite-to-ferrite phase transformation was almost completed in both steels after the shorter DT at temperatures of 630 and 680 °C, although a harder microstructural constituent (bainite) was occasionally observed in these specimens after the CT of 300 s (Figure 2a).
The EBSD analyses confirmed that a fully ferritic matrix formed in both steels after the 1800 s DT at every CT.The results also showed that the quasipolygonal ferrite grains observed by Kim et al. [20] and Cheng et al. [28] in their hot-deformed steels held isothermally at temperatures of 570 and 600 °C; 620 °C was detected occasionally in the Ti-Mo-V steel only at the temperature of 590 °C, as shown in Figure 2c,d.
Table 2 presents the summary of the average grain size measured by EBSD in these steels after the simulated TMCP process and different coiling parameters.Figure 3 presents examples of the EBSD analyses, showing ferrite grains separated by the grain boundaries with misorientation higher than 10°and also the corresponding grain size distributions.Some lath-type bainitic structure is also evident in the grain size image after coiling at 590 °C for 300 s, as shown in Figure 3a.It must be noted that the average grain sizes measured by the equivalent circle diameter (ECD) method do not necessarily show the grain size variations (ratio of small and large grains), meaning that samples might have the same average grain size but different grain size distribution (homogeneity).
It should be noted that the grain size of the low-carbon and low-alloyed steels depends on the TMCP process and the subsequent cooling procedures.[34] However, some exceptions were found in the literature; fine (<3 μm) ferrite grains were formed in Ti-Mo-Cr-(V) steels during continuous cooling after the 50% and 30% hot deformation at temperatures of 1050 and 900 °C, respectively, whereas 20% hot deformation at the temperature 900 °C, followed by the subsequent isothermal dwell at the temperature of 650 °C for 1 h, resulted in the relatively large ferrite grain size of 30 μm in a 0.1Ti-0.2Mosteel. [26,35]In the former case, the enhanced hardenability of the steels due to the extra Cr alloying might cause rapid nucleation and growth of allotriomorphic ferrite grains at the intermediate temperatures during the slow cooling (cooling rate ≦1 °C s À1 ).In the latter case, it was reported that the 0.1Ti-0.2Mosteel contained nearly 50% austenite after 5 min at the isothermal temperature of 650 °C.This sluggish austenite-to-ferrite transformation kinetics due to the relatively low, 20% deformation probably allowed the allotriomorphic ferrite grains to grow relatively large before they were impinged by other ferrite grains.A similarly sluggish austenite-to-ferrite transformation kinetics was observed in the specimens of this study aged at the temperature of 590 °C, although their grain size was less than 3 μm.This can be explained by the notably higher true strain of the hot compression under the T nr temperature of the test steels.This resulted in pancake austenite grains with high dislocation density which, in turn, provide numerous potential nucleation sites for ferrite grains with higher undercooling compared to the hot-deformed steel with the 20% deformation isothermally aged at the temperature of 650 °C.
Table 3 and 4 present the summary of the precipitates observed in these steels after the simulated TMCP process and different coiling parameters.The estimated fraction of the Table 2. Average grain size (>10°) of Ti-Mo-V and Ti-Mo-V-Nb steels after simulated TMCP and different coiling parameters.carbides is based on assumptions that particles detected in a certain STEM image area were spherical and their average size was also their radius.Figure 4 presents examples of typical precipitates formed in the steels.The thickness of the thin foils was measured with the Transmission electron microscopy-Electron energy loss spectroscopy (TEM-EELS) Log-Ratio technique used in ref. [36].The average thickness (l) of the present specimens in the regions of interest was about 50 nm (ranging from 40 to 60 nm).With the number of particles counted in each STEM image (image area = h Â k), the average fraction of precipitates was estimated in a certain volume (h Â k Â l = volume in m 3 ) of the ferrite matrix.Otherwise, ref. [27] presents detailed analyses of the precipitation behavior of the test steels.Only a few main observations from Table 3 and 4 are highlighted for the reader because the precipitation behavior of the test steels was reported in detail elsewhere. [27]The most interesting observation was the absence of the IPs after the isothermal aging at temperatures of 590 and 630 °C in the Ti-Mo-V steel.Similar precipitation behavior was reported earlier for 0.1Ti steel and 0.2Ti-0.2Mosteel. [29,37]It was also shown earlier that unstable IPs formed during the phase transformation in the 0.1Ti steels, but they dissolved during the increased aging time. [29]egarding the Ti-Mo-V-Nb steel, the Ti and Nb alloying had no significant effect on the formation of the IP because they formed large (Ti,Nb)Cs during the hot deformation.Thus, the nanosized IPs were precipitated at temperatures of 630 and 680 °C during the phase transformation.These carbides had an exact Baker-Nutting orientation relationship with the polygonal ferrite matrix at the CT of 680 °C, whereas a near Baker-Nutting orientation relationship was observed in the specimen isothermally aged at the temperature of 630 °C.However, the formation of these IPs was restricted at the temperature of 590 °C.The spherical vanadium-rich carbides that had a near Baker-Nutting orientation relationship with the polygonal ferrite matrix formed in the dislocation walls of the supersaturated ferrite.

Estimation of the Yield and TS
In accordance with the earlier studies, [38][39][40] the hardness of the specimens was measured, and their estimated values of the YS and the TS were calculated using the Equation ( 1) and ( 2). [41] ¼ À90.7 þ 2.876 HV (1) Table 3.Average size [27] and volume fraction of different Ti(V,Mo)C carbide types in Ti-Mo-V steel after 1800 s. (IP = interphase precipitates).Table 4. Average size [27] and volume fraction of different V(Mo)C carbide types in Ti-Mo-V-Nb steel after 1800 s. (IP = interphase precipitates).

Steel and process parameters
Ti-Mo-V-Nb Results are presented in Table 5. Considering the standard error of 112 MPa, which is related to the value given by Equation (2), the TS values correspond well with the values given for the low-alloyed steels in the international standard ISO 18265. [42]Thus, it was predicted that the TS of 1 GPa could be achieved in the steels using the present TMCP with the CTs of 590 and 630 °C.

Dislocation Density
The well-known Williamson-Hall method has generally been used to estimate the dislocation density of different types of steels.The method unfortunately overestimates the dislocation density of the quenched martensitic steels.The overestimation is caused by the strong elastic anisotropy of the martensite, which induces a nonlinear behavior of the magnitude of the diffraction vector (K ) and the broadening of the diffraction peaks (ΔK ).The modified Williamson-Hall equation with the contrast factor C and/or modified Warren-Averbach analysis, presented by Ungár et al., [43] is often applied in these cases to evaluate the dislocation density.
However, Takebayashi et al. [44] have shown that the tempering at the 450-650 °C temperature range reduced the elastic anisotropy of the quenched steels.The tetragonal martensitic crystal structure with the dense crystal defect distribution was transformed to the ferritic cubic crystal structure during the tempering, so the dislocation densities evaluated by using Williamson-Hall method gave comparable values in the studied steels.[47][48][49][50][51][52] No martensitic microstructures with the high dislocation density exist in the present steels, and their microstructure is comparable with the tempered steel consisting of fully ferritic matrix-containing precipitates; thus, this study calculated the dislocation densities using the classic Williamson-Hall method (Equation (3)) where ρ s is the dislocation density calculated from strain broadening and ρ p is the dislocation density calculated from crystallite size.The following Equation ( 4) and ( 5) were used to evaluate ρ s and ρ p of the body-centered cubic metal ρ s ¼ 14.4ε 2 =Fb 2 (4) where ε is microstrain, b is Burgers vector, F is an interaction factor assumed to be 1, and D is crystallite size.In the Equation ( 5), n is dislocations per block face and its value is assumed to be 1 because this value will lead to the minimum dislocation density, as ref.[51] explained earlier.
Table 6 shows the microstrain, crystallite size, and dislocation densities of the specimens.The value of Burgers vector was   2.48 Å.The specimens coded 590-300 were not analyzed due to the retained austenite that affected the ferrite (100) diffraction peak's shape.
The results showed that the dislocation density increased with the decreasing CTs.The dislocation density in both steels was notably higher at all CTs after the isothermal DT of 1800 s compared to the typical estimated value of 5 Â 10 13 m À2 used in the literature. [15,32,33]A notable increase of the dislocation density was also observed in the specimens isothermally transformed at the 590 °C temperature compared to the other CTs.These values were comparable with the low carbon Mo-V-Ti-N steels with a mixed microstructure of polygonal ferrite, acicular ferrite, granular bainitic ferrite, and a martensite-austenite constituent, [53] although the present steels were fully ferritic, as shown in Figure 1.Thus, the present results demonstrated that a significant difference between the dislocation densities can be achieved in the low-carbon low-alloyed steels by changing the CT.

Contribution of Strengthening Mechanisms to YS
Five strengthening mechanisms can generally contribute to the YS of a steel.These mechanisms are lattice friction stress of ferrite matrix σ 0 , solid solution strengthening σ ss , grain-boundary strengthening σ g , precipitation strengthening σ g , and dislocation strengthening σ d .Their effects are generally considered independent factors, as seen in Equation ( 6), although a simple addition of the strengthening from dislocations and precipitates may result in an overestimation of their contribution, as pointed out in the literature. [32,33,48] The first factor of Equation ( 6) is the lattice friction stress of ferrite.Regarding the low-carbon, low-alloyed ferritic steels, this value of σ 0 varies from 45 to 54 MPa in the literature. [15,19,20,32,33]his study used the value of 45 MPa.
The second factor is the solid solution hardening due to solute elements.It is calculated using Equation ( 7) Reference [15] used Equation (7), and it was also applied in this study, although the equation does not include the main carbide forming elements (Ti, Mo, V, and Nb).This study also adopted the assumptions concerning the carbon and nitrogen content made in ref. [15].Thus, the estimated carbon and nitrogen content were 0.01 and 0 wt%, respectively.According to Equation ( 7), the estimated value of solid solution strengthening in both steels of the present study is 129 MPa.Comparing the result to the other studies, it obviously matches the value of 125 MPa presented by Huang et al., [15] but it is notably lower than the estimated value of 202 MPa of Kim et al. [20] It should be noted that Kim et al. [20] included Ti and Cr content in their calculations, but they did not give any estimation of the amount of free carbon and nitrogen atoms.However, the present value is higher than the estimated values of 102 and 49 MPa in the MoCrNbV and VCrNb steels, respectively. [19]These values were based on the Atom probe tomography (APT) measurements, which naturally gave a more accurate estimation of the solid solution strengthening compared to the assumptions related to the free interstitial content this study used.The assumption of the carbon content of 0.01 wt% matches the measured carbon content of the MoCrNbV steel, although Singh et al. [19] did not present any value for nitrogen content.However, the measured carbon content of the VCrNb steel was as low as 0.001 wt% in their study.This explained the minor contribution of the solid solution strengthening to the YS of this steel, although the factor of the combined C and N content in the equation used in ref. [19] was higher than the factor used in the present study.Therefore, it was concluded that the present calculated value of 129 MPa might be a slight overestimation because the exact amount of free interstitial atoms and solute atoms could not be determined in the test steels within different coiling parameters in the cases of the test steels isothermally aged at the temperatures.The most accurate estimation of the solid solution strengthening in steels would obviously require APT measurements and a reliable regression equation that included all the major alloying elements.
A third factor determines the effect of the grain-boundary strengthening on the YS.It is calculated using the well-known Hall-Petch relation (Equation ( 8)) where k is the grain boundary resistance to the dislocation movement.The value of the constant k varies slightly in literature; [15,19,20,48] nevertheless, the value of 18.1 MPa mm À0.5 used in ref. [19] was applied in this study.Huang et al. [15] reported that the contribution of grain size strengthening depended on the CTs (600-700 °C), and the average grain size, varying from 3.5 to 6.2 μm, corresponded to the calculated YS values 304 and 210 MPa, respectively.Table 7 and 8 Table 6.Microstrain, crystallite size, and dislocation density of Ti-Mo-V and Ti-Mo-V-Nb steels.show that the results of the present studies were quite comparable to those YS values.However, it was reported that the average grain size (3 μm) did not vary despite the large CT range of 570-670 °C after the TMCP, in which the FRT of 880 °C was applied.In this case, the contribution to the YS of the steel was 318 MPa. [20]he contribution of the dislocation hardening to the yield stress can be estimated using Equation ( 9) where α is the constant (0.3), M is average Taylor factor 2.75, G is the shear modulus of ferrite (81 600 MPa), b is the Burgers vector, and ρ is the dislocation density.Table 7 and 8 present the contribution of the dislocation strengthening to the YS of the test steels using the data shown in Table 6.][36] The latter value was strictly consistent with the value of 169 MPa, which was based on the TEM studies reported by Singh et al. [19] However, Kim et al. [20] suggested that the dislocation strengthening could be neglected due to its minor contribution to the YS of their steels.The present results indicated that the contribution of the different strengthening mechanisms depended strongly on the CT and steel chemistry.The increased dislocation strengthening especially compensated the adverse effect of the reduced precipitation strengthening on the YS of the steel when the CT was too low for the IP formation, even in the Ti-Mo-V-Nb steel.
The amount of the precipitation hardening σ p is usually evaluated using the Ashby-Orowan relationship as follows Δσ p ¼ ðK=dÞ f 0.5 lnðd=bÞ (10)   where K is the constant with the value of 5.9 N m À1 , f is the volume fraction of precipitates (%), d is the average diameter of precipitates (μm), and b is the Burgers vector.However, a reliable estimation of the precipitation hardening in the test steels is very difficult because the precipitate type and size could vary within a single ferrite grain.This phenomenon has also been reported in the earlier studies. [35,54,55]Therefore, the effect of precipitation hardening on the YS was calculated using Equation (11), which was based on the assumption that the increment of precipitation hardening can be calculated by subtracting friction stress of pure iron, solid solution strengthening, grain boundaries strengthening, and dislocation strengthening from total YS. [31] p ¼ σ y À Δσ 0 À Δσ ss À Δσ g À Δσ d (11)   In this case, the value of σ y was taken from the calculated values presented in Table 4.The estimated contribution of the precipitation hardening to the YS is shown in Table 7 and 8.
The results showed that the precipitation hardening was negligible at the CT of 590 °C, whereas the estimated values of the Ti-Mo-V-Nb steel isothermally aged at the temperature of 630 °C were highly in consistent with the results reported by several studies, [20,33,34] although a stronger effect of the precipitation hardening on the yield was also reported in the literature. [15,19]he small contribution of the precipitate strengthening of the specimen isothermally aged at the 680 °C temperature may be due to the relatively large size of the IPs and spherical carbides and their smaller volume fraction compared to the specimen isothermally aged at the 630 °C temperature.
Regarding the Ti-Mo-V steel, the predicted strength increments due to the precipitation strengthening after the aging at temperatures of 630 and 680 °C seem reasonable because the volume fractions and the sizes of the precipitates correspond with the results shown in Table 1.The negligible contribution of the relatively large spherical carbides to the YS at the temperature of 590 °C is a quite reasonable estimation and comparable with the precipitation behavior of the Ti-Mo-V-Nb steel.
Finally, comparing the contributions of the precipitation strengthening and the dislocation strengthening, it was concluded that increased dislocation density can compensate the decreased effect of the precipitation hardening (Figure 5).
This can explain the high (around 900 MPa) YS of a fully ferritic steel without a fine, evenly distributed precipitation structure.Thus, using the present simulated TMCP process, a fully ferritic microstructure and the UTS value of 1 GPa were achieved in both steels despite the different steel chemistries and precipitation behavior.The mechanical properties and the stretch flangeability of these steels should be determined in the near future in order to evaluate their potential as a new, ultrahigh-strength steel grade for the automotive industry.

Conclusions
The main observations and conclusions of this study can be summarized as follows: A fully ferrite matrix can be achieved in both test steels using the present TMCP and the 30 min coiling time at every tested CT.The aimed TS of 1 GPa could be achieved in them using the CTs 590 and 630 °C.Therefore, a CT range of 590-630 °C with the 1800 s DT seemed to be a potential process window for the upcoming laboratory scale hot-rolling experiments.A fine average polygonal ferrite grain size, 3-5 μm, can be achieved in both steels with the present TMCP process and CTs.Thus, the contribution of the grain boundary strengthening to the YS can be from 290 to 340 MPa in the test steels.
The dislocation density increased with the decreasing CTs.Its effect on strengthening of the steels was significantly higher at all CTs compared to the typical value of 5 Â 10 13 used in the literature.Moreover, the results showed that decreasing the CT can notably increase its effect on the YS, and they also indicated that the increased dislocation density can compensate the decreased effect of the precipitation hardening on the YS in fully ferritic 1 GPa low carbon low-alloyed steels.
The slow nucleation rate of the spherical Ti(Mo,V) carbides in the Ti-Mo-V steel resulted in coarser precipitates compared to the spherical V(Mo)C precipitates in the Ti-Mo-V-Nb steel, although the relative phase fraction of these carbides was identical in the test steels after the isothermal aging at temperatures of 630 and 680 °C.
The relative fraction of the spherical carbides formed after the phase transformation in the Ti-Mo-V-Nb steel was comparable with the fraction of the spherical carbides formed in the Ti-Mo-V steel at the lowest CT of 590 °C.However, the low alloy content within the supersaturated ferrite matrix caused by the formation of the interphase precipitates containing V and Mo during the phase transformation and numerous, potential nucleation sites (i.e., dense dislocation structure) resulted in the formation of nanosized spherical V(Mo) carbides with a smaller particle size compared to the carbides in the Ti-Mo-V steel.Nevertheless, the contribution of the precipitation strengthening was negligible at this CT in both steels due to the low volume fraction of these carbides and their relatively large size.

Figure 1 .
Figure 1.Dilatation curves of the specimens: a) Ti-Mo-V steel and b) Ti-Mo-V-Nb.The images are based on data presented in ref. [27].

Figure 2 .
Figure 2. Microstructures of a) Ti-Mo-V steel and b) Ti-Mo-V-Nb steel after being isothermally transformed at 630 °C for 300 s (bainite marked as red arrows).EBSD imaging showing high-and low-angle boundaries in ferritic matrix of c) Ti-Mo-V steel and d) Ti-Mo-V-Nb steel after being isothermally transformed at 590 °C for 1800 s.Black and red lines indicate the misorientation angle θ ≥ 10°and 2°≤ θ < 10°, respectively.

Figure 4 .
Figure 4. Typical precipitates of a) Ti-Mo-V steel and b) Ti-Mo-V-Nb steel after being isothermally transformed at 630 °C for 300 s.

Table 1 .
Chemical composition of the test steel alloys (wt%).

Table 8 .
The combined contribution of strengthening mechanisms to the estimated YS σ y of the Ti-Mo-V-Nb steel after the CT of 1800 s calculated using Equation(6).The calculated YS is given in parentheses.

Table 7 .
The combined contribution of strengthening mechanisms to the estimated YS σ y of the Ti-Mo-V steel after the CT of 1800 s calculated using Equation(6).The calculated YS is given in parentheses.