Scientist

Dirk Ponge
Group Leader
Phone: +49 211 6792 438

Publication Reference

1.
D. Raabe, D. Ponge, O. Dmitrieva, and B. Sander, "Designing Ultrahigh Strength Steels with Good Ductility by Combining Transformation Induced Plasticity and Martensite Aging," Advanced Engineering Materials 11, 547-555 (2009).

Research Topics - Advanced High Strength and Nanostructured Steels

Designing Ultrahigh Strength Steels with Good Ductility by Combining Transformation Induced Plasticity and Martensite Aging

 Steels with a high ultimate tensile strength (UTS) above 1 GPa and good ductility (total elongation (TE) of 15-20% in a tensile test) are of greatest relevance for lightweight engineering design strategies and corresponding CO2 savings. In this project we work on a novel design approach for precipitation hardened ductile high strength martensitic and austenitic-martensitic steels (up to 1.5 GPa strength).

Dierk Raabe, Dirk Ponge, Olga Dmitrieva, Benedikt Sander

 Abstract

Steels with a high ultimate tensile strength (UTS) above 1 GPa and good ductility (total elongation (TE) of 15-20% in a tensile test) are of greatest relevance for lightweight engineering design strategies and corresponding CO2 savings.

In this project we work on a novel design approach for precipitation hardened ductile high strength martensitic and austenitic-martensitic steels (up to 1.5 GPa strength). The alloys are characterized by a low carbon content (0.01 wt.% C), 9-15 wt.% Mn to obtain different levels of austenite stability, and minor additions of Ni, Ti, and Mo (1-2 wt.%). The latter are required for creating precipitates during aging heat treatment.

Steels with a high ultimate tensile strength (UTS) above 1 GPa and good ductility (total elongation (TE) of 15-20% in a tensile test) are of paramount relevance for lightweight engineering design strategies and corresponding CO2 savings.

Figure 1: Overview of the typical strength-ductility profiles of different types of steels. The strength is expressed in terms of the ultimate tensile strength measured during tensile testing and the ductility is expressed in terms of the total sample elongation. The data represent regimes such as published in the references given below. TRIP: transformation-induced plasticity; TWIP: twinning-induced plasticity; Complex phase: multiphase steels (e.g. austenitic-martensitic steels which may contain also bainite); maraging TRIP: new steel concept that includes hardening mechanisms based on transformation induced plasticity and the formation of intermetallic nanoparticles in the martensite during aging. The approach leads to an unexpected simultaneous increase in both strength and total elongation (green area) enhancing the regime of formable ultrahigh strength steels by 0.5 GPa. Zoom Image

Figure 1: Overview of the typical strength-ductility profiles of different types of steels. The strength is expressed in terms of the ultimate tensile strength measured during tensile testing and the ductility is expressed in terms of the total sample elongation. The data represent regimes such as published in the references given below. TRIP: transformation-induced plasticity; TWIP: twinning-induced plasticity; Complex phase: multiphase steels (e.g. austenitic-martensitic steels which may contain also bainite); maraging TRIP: new steel concept that includes hardening mechanisms based on transformation induced plasticity and the formation of intermetallic nanoparticles in the martensite during aging. The approach leads to an unexpected simultaneous increase in both strength and total elongation (green area) enhancing the regime of formable ultrahigh strength steels by 0.5 GPa.

[less]

Here we report about our novel design approach for precipitation hardened ductile high strength martensitic and austenitic-martensitic steels (up to 1.5 GPa strength). The alloys are characterized by a low carbon content (0.01 wt.% C), 9-15 wt.% Mn to obtain different levels of austenite stability, and minor additions of Ni, Ti, and Mo (1-2 wt.%). The latter are required for creating precipitates during aging heat treatment.

Hardening in these materials is realized by combining the TRIP effect with a maraging treatment (TRIP: transformation-induced plasticity; maraging: martensite aging through thermally stimulated precipitation of particles). The TRIP mechanism is based on the deformation-stimulated athermal transformation of metastable austenite (face centered cubic Fe-Mn phase) into martensite (metastable body centered cubic or orthorhombic Fe-Mn phase) and the resulting matrix and martensite plasticity required to accommodate the transformation misfit. The maraging treatment is based on hardening the heavily strained martensite through the formation of small intermetallic precipitates (of the order of several nanometers). These particles act as highly efficient obstacles against dislocation motion through the Orowan and Fine-Kelly mechanisms enhancing the strength of the material.

While both types of alloys, i.e. TRIP steels  and maraging steels have been well investigated, the combination of the two mechanisms in the form of a set of simple Fe-Mn alloys as suggested in this work, namely, the precipitation hardening of transformation-induced martensite by intermetallic nanoparticles, opens a novel and lean alloy path to the development of ultrahigh strength steels that has not been much explored in the past.

Materials, Experiments, and Characterization

Four alloys were investigated, namely, one standard 17 wt.% Ni-11 wt. Co maraging steel and the three maraging TRIP Fe-Mn steels (9 wt.% Mn, 12 wt.% Mn, 15 wt.% Mn).[12,13]
The three Fe-Mn alloys have a low carbon content and minor additions of Ni, Ti, Al, and Mo to form precipitates in the martensite.[36] The main difference among the three materials consists in their Mn content (~9 wt.%, ~12 wt.%, ~15 wt.%) and hence, in the volume fraction and stability of the retained austenite they contain after quenching.

Table 1: Alloy composition in weight % (wt%) Zoom Image
Table 1: Alloy composition in weight % (wt%)

The alloys were melted and cast to round billets of 1kg each in a vacuum induction furnace. Annealing and swaging of the as-cast alloys was conducted to ensure homogenization of the microstructure and removal of segregation effects. After annealing at 1150°C for 1h swaging was conducted in 8 passes between 1000°C and 1150°C. The billets were swaged from a diameter of 27.0 mm to 13.5 mm which corresponds to a logarithmic strain of 1.39. This was followed by air cooling to 800°C and a water quench to room temperature. The rods were reheated to 1100°C for 0.5h, hot rolled in 6 passes to a total logarithmic strain of 1.9 into strips with a thickness of 4 mm and water quenched. These strips were cold rolled to a thickness of 1.5 mm corresponding to a logarithmic strain of 1. The subsequent solution heat treatment was performed at 1050°C for 0.5h followed by a final water quench. For heat treatments above 1000°C Argon gas atmosphere was used to prevent oxidation. Final aging heat treatments were conducted at different temperatures between 425°C and 500°C at times between 1 minute and 48 hours.
Flat tensile specimens were machined in the as quenched and in the solution treated state with a thickness of 1mm, width of 4mm and a gage length of 10 mm. A strain gage extensometer was used for precise determination of the strain. Tensile testing was conducted on a Zwick ZH 100 tensile testing machine at a constant cross head velocity corresponding to an initial strain rate of 8 x 10-1 s-1.

Hardness testing according to EN ISO 6507-1 was conducted using a Zwick 3212 hardness tester to determine the Vickers hardness with a load of 49.05 N (HV5).
Characterization of the chemical and microstructural homogeneity of the cast, formed, and heat treated samples was conducted by using optical and scanning electron microscopy (SEM) in conjunction with EDX (energy dispersive x-ray spectrometry) and high resolution EBSD (electron back scatter diffraction). The SEM was a JEOL JSM-6500F field emission scanning electron microscope (FE-SEM) operated at 15 kV. The EBSD scans were carried out in areas of about 100 x 270 mm2 in cross sections in the middle of the samples at a step size of 500 nm. Samples were ground using SiC paper (8 µm). Subsequently, the samples were mechanically polished using diamond suspensions of 3µm and 1µm. Final polishing was done using a SiO2 suspension (0.1µm). In order to study the possible influence of mechanical polishing on premature transformation of retained austenite (which is metastable against shear loads) samples were also prepared by electropolishing using 500ml methanol, 500ml 2-butoxyethanol and 60ml 70% perchloric acid for 20 seconds. Both preparation methods provided similar microstructure results.
Transmission electron microscopy (TEM) images were taken on the solution-treated, quenched plus finally age hardened sample with 12 wt.% Mn in order to study the size and spatial distribution of the nanoparticles which are formed during aging. For TEM sample preparation the material was first thinned to a thickness below 100 µm by mechanical polishing. Standard 3-mm TEM discs were then punched and electropolished into TEM thin foils using a Struers Tenupol twin-jet electropolishing device. The electrolyte consisted of 5% perchloric acid (HClO4) in 95% ethanol cooled to −30°C. The thinned specimens were then investigated in the field emission transmission electron microscope JEOL JEM 2200 FS operated at 200 kV. The analysis was carried out in scanning TEM mode (STEM) using a bright field (BF) detector.

The chemical composition of some of the nanoparticles observed in the material was studied at atomic scale resolution using atom probe tomography (APT) with a local electrode technique (IMAGO LEAP 3000X HR metrology device). This system provides an excellent mass resolution which is essential for the analysis of multi-component steels. The configuration includes a laser based atom probe capability in addition to voltage pulsing. The measurements presented in this work were conducted in laser mode where the LEAP electrode applies a static field to the specimen while an ultra-fast laser pulse triggers the removal of the atoms. The use of the reflectron method used in the current configuration is important for an optimal time of flight (TOF) precision. Sample tip preparation for APT analysis was conducted using perchloric acid chemical etching.

Results

Figure 2: Engineering stress–strain curves for the four steels, Table 1. (a) Standard 17wt% Ni-11 wt% Co maraging steel; (b) Fe–Mn steel with 9wt% Mn; (c) Fe–Mn steel with
12 wt% Mn; (d) Fe–Mn steel with 15 wt% Mn. All alloys have a very low carbon content and minor additions of Ni, Ti, and Mo to form precipitates. The main difference among the
Fe–Mn alloys consists in their Mn content and hence, in the amount of retained austenite they contain. Three sets of data are shown, namely, in the as-quenched, age hardened, and
15% cold rolled plus age hardened state. Zoom Image
Figure 2: Engineering stress–strain curves for the four steels, Table 1. (a) Standard 17wt% Ni-11 wt% Co maraging steel; (b) Fe–Mn steel with 9wt% Mn; (c) Fe–Mn steel with
12 wt% Mn; (d) Fe–Mn steel with 15 wt% Mn. All alloys have a very low carbon content and minor additions of Ni, Ti, and Mo to form precipitates. The main difference among the
Fe–Mn alloys consists in their Mn content and hence, in the amount of retained austenite they contain. Three sets of data are shown, namely, in the as-quenched, age hardened, and
15% cold rolled plus age hardened state.
[less]

Fig. 2 shows the mechanical data of the four alloys. Fig. 3 presents the corresponding microstructure results (top row: phase distribution of α’-martensite and of retained austenite; bottom row: microtexture in terms of the {hkl} Miller triple and high angle grain boundaries). Fig. 2a shows the engineering stress-strain curves of the conventional Ni-Co maraging steel in as-quenched, age hardened, and 15% cold rolled plus age hardened state. The latter state was studied as the mechanical properties after pre-deformation prior to the tensile test may indicate the occurrence of a TRIP effect. The corresponding results obtained from phase and texture determination via EBSD are presented in Fig. 3a, confirming that the Ni-Co maraging steel contains no or little retained austenite in the as-quenched state.

The yield strength (YS) of the material with 9 wt.% Mn is about 350 MPa, its ultimate tensile strength (UTS) about 810 MPa, and the total elongation (TE) about 6% in the as-quenched state. The properties after aging heat treatment (48 hours at 450°C) are surprising. The UTS lies above 1 GPa (as expected) while the TE does not drop upon precipitation strengthening as observed for conventional Ni-Co based maraging steels[12-16] but it increases from 6% to more than 15%. This means that precipitation hardening in this martensitic alloy simultaneously increases both strength and ductility. 15% pre-deformation (by cold rolling) of the same sample prior to aging yields similar properties as without pre-deformation. This results indicates that no retained martensite is involved. This is confirmed by the EBSD maps. The EBSD analysis for the 9 wt.% Mn alloy (Fig. 3b) reveals coarse α’-martensite lamellae with longitudinal dimensions of up to 100 mm, but no retained austenite appears in the solution annealed and quenched state. This means that it is not an age hardenable TRIP steel (referred here to as maraging TRIP steel) but a Mn-based maraging steel.[18,19]

Similar observations are made for the microstructure of the 12 wt.% Mn alloy, Figs. 2c, 3c. This sample also consists of an α’-martensite matrix but it contains up to 15 vol.% retained austenite and some ε-martensite. Hence, this material represents an age hardenable TRIP steel as it can undergo hardening both, via mechanically induced martensite formation and also through precipitation hardening of the as-quenched and of the mechanically induced martensite. The EBSD map for the 12 wt.% Mn alloy shows a considerably finer α’-martensite microstructure than that observed in the 9 wt.% Mn sample, Figs. 3b,c. Dilatometry (measurement of thermal expansion) and ferromagnetic data suggest in part a higher austenite fraction of up to 20 vol.%. The differences between dilatometry, ferromagnetic characterization, and EBSD analysis can be attributed to the limited statistics provided by EBSD. Also EBSD yields surface information only. The size of the retained austenite islands lies between 1 mm and 20 mm. The ε-martensite lamellae are below 2 mm and occupy an overall fraction of about 1-2 vol.%.
The 12 wt.% Mn alloy has a YS of about 325 MPa, UTS of nearly 1 GPa, and TE of about 16% in the as-quenched state, Fig. 2c. After aging (450°C for 48 hours) the UTS increases to more than 1.3 GPa and the TE to 21%. 15% cold rolling prior to aging leads to a strong increase in strength (nearly 1.5 GPa UTS) but also to a drop in TE (about 10%).

Figure 3: Microstructure results for the four alloys obtained via EBSD analysis (100 nm step size). Top row: phase
distribution of a0-martensite (green) and of retained austenite (red). Bottom row: microtexture in terms of the
{hkl} Miller triple color coding and high angle grain boundaries (black, >15 8). (a) Conventional 17 wt%
Ni-11 wt% Co maraging steel; (b) Fe–Mn steel with 9wt% Mn; c) Fe–Mn steel with 12 wt% Mn; (d) Fe–Mn
steel with 15 wt% Mn, Table 1. Zoom Image
Figure 3: Microstructure results for the four alloys obtained via EBSD analysis (100 nm step size). Top row: phase
distribution of a0-martensite (green) and of retained austenite (red). Bottom row: microtexture in terms of the
{hkl} Miller triple color coding and high angle grain boundaries (black, >15 8). (a) Conventional 17 wt%
Ni-11 wt% Co maraging steel; (b) Fe–Mn steel with 9wt% Mn; c) Fe–Mn steel with 12 wt% Mn; (d) Fe–Mn
steel with 15 wt% Mn, Table 1.
[less]

Regarding the ductility it is most remarkable that both steels (9 wt.% Mn and 12 wt.% Mn) show, irrespective of their retained austenite content, the surprising feature of a simultaneous increase in both, UTS and total elongation upon aging. While the UTS increases by 25-30% the total elongation increases by more than 150% (from 6% to 15%) for the 9 wt% Mn sample and by 31% (from 16% to 21%) for the 12 wt.% Mn alloy. This increase of both properties represents a very unusual feature of these ultrahigh strength materials.

Fig. 2d shows the engineering stress-strain curves for the 15 wt% Mn sample steel in as-quenched, age hardened, and 15% cold rolled plus age hardened state. The corresponding phase and texture maps are given in Fig. 3d. The alloy contains a high volume fraction of retained austenite in the as-quenched state. The high resolution EBSD analysis presented in Fig. 4 shows that in addition to α’-martensite the sample also contains some ε-martensite (marked in blue).[37]

The as-quenched 15 wt.% Mn alloy has a YS of about 160 MPa, a UTS of about 800 MPa, and a TE of about 40%, Fig. 2d. After aging (450°C for 48 hours) the UTS drops to about 700 MPa and the TE to 32%. 15% cold rolling prior to aging leads to a strong increase in strength (above 1.2 GPa UTS) but also to a drop in TE (about 3%).
This means that the steel with the most stable austenite and highest Mn content (15 wt.%) does not reveal the same unusual feature of an increase in ductility upon aging heat treatment as the two steels with smaller Mn content (9 wt.% and 12 wt.%).

The BF-STEM micrographs that were taken exemplarily on the age hardened maraging TRIP steel (12 wt.% Mn) show that the nanoscaled precipitates reveal a narrow size distribution with an average diameter of 8-12 nm, Figs. 5a,b. Fig. 5b also reveals some of the retained austenite which is free of precipitates.

Figure 4: High resolution EBSD analysis of the 15 wt% Mn steel showing details of the differentiation between
retained austenite (red), a0-martensite (green), e-martensite (blue) (50 nm step size). Zoom Image
Figure 4: High resolution EBSD analysis of the 15 wt% Mn steel showing details of the differentiation between
retained austenite (red), a0-martensite (green), e-martensite (blue) (50 nm step size).
[less]

Local EDX analysis conducted in convergent beam mode show an increased content in Ni, Ti, and Al in these particles when compared to the surrounding matrix.[13,14,38] The area density of the nanoparticles was about 300 µm-2. The corresponding volume density is estimated as 5×103 µm-3 (about 2-3 vol. %), Fig. 5. Slightly elongated nanoparticles were observed at the interfaces between the martensite lamellae.
In addition to the EDX-TEM analysis we also conducted APT measurements on the alloy with 9 wt.% Mn. Fig. 6 reveals a somewhat more complex composition of the precipitates. The APT maps show the joint occurrence of Ni, Al, and Ti in some of the clusters and also some isolated arrays where either Ni or Ti coincide with higher concentrations of Al. A more detailed atomic scale analysis inside some of these particles revealed a composition including not only Ni, Al, and Ti but also Fe and Mn enrichment. These observations indicate that the particles encountered do as a rule not assume a simple binary and ternary composition but seem to be more complex. In some case even the formation of tiny austenitic clusters or zones similar to the FeMn q phase is conceivable.

Besides the nanoparticles also larger particles were observed in the Mn-alloyed samples via TEM, e.g. Fig. 5. They had an average size of 1-2µm. EDX revealed that these larger precipitates contained Ti, Mo, C, and Al which did not allow for a closer estimation of their constitution.

Discussion

The stress-strain data for the different specimens show in principal two types of behavior before aging. The first two samples, i.e. the conventional highly alloyed 17 wt.% Ni-11 wt.% Co maraging steel and the maraging steel with 9 wt.% Mn both reveal an α’-martensitic microstructure in the as-quenched state, Fig. 3. This means that both materials do not contain metastable retained austenite, hence no TRIP effect occurs. Their mechanical performance, observed in tensile tests, Figs. 2a,b, reflects this fact. Both materials reveal the same strength-elongation profile in the as-quenched and in the 15% pre-rolled state, i.e. no TRIP-related hardening was observed.
In contrast, the 12 wt.% Mn and the 15 wt.% Mn samples, table 1, both show a pronounced increase in strength upon pre-deformation (plus subsequent aging), Figs. 2, 3. This observation is attributed to the mechanically-stimulated transformation of the retained austenite leading to a TRIP effect. Corresponding EBSD measurements show that the retained austenite present in both specimens in the as-quenched state is gradually transformed into martensite during deformation, Fig. 7. The retained austenite is also present after heat treatment as the aging temperature (450°C) was chosen below the temperature for re-transformation from martensite back into austenite. This point will be discussed in more detail below. Fig. 7 shows EBSD data that were taken on the 12 wt.% Mn sample in the as-quenched plus aged state, Fig. 7a, and the same microstructure at uniform elongation in the deformed flat tensile specimen, Fig. 7b.
The motivation for attributing the additional strain hardening capacity of the 12 wt.% and 15 wt.% Mn alloys essentially to the TRIP effect becomes clear when comparing the data for the 15% pre-rolled specimens among the four samples, Fig. 2. While the samples with 12 wt.% Mn and 15 wt.% Mn contain retained austenite, Fig. 5b, and, hence, show a strong increase in strength upon pre-deformation, the conventional Ni-Co maraging steel and the 9 wt.% Mn samples both do not contain retained austenite and, therefore, do not reveal any change in strength when 15% cold rolled prior to tensile testing.

Figure 5: TEM images of nanoparticles (and some larger particles) formed in the aged 12 wt% Mn alloy (450 8C, 48 h). The nanoparticles have an average diameter of 8–12 nm. Local EDX analysis shows an increased content in Ni, Ti, and Al in the
particles relative to the matrix. (a) BF-STEM images of particles in two neighboringmartensite lamellae. (b) BF-STEM images of particles in the martensite and particle-freeaustenite regions. Zoom Image
Figure 5: TEM images of nanoparticles (and some larger particles) formed in the aged 12 wt% Mn alloy (450 8C, 48 h). The nanoparticles have an average diameter of 8–12 nm. Local EDX analysis shows an increased content in Ni, Ti, and Al in the
particles relative to the matrix. (a) BF-STEM images of particles in two neighboringmartensite lamellae. (b) BF-STEM images of particles in the martensite and particle-freeaustenite regions.
[less]

The second approach to classify the mechanical test result lies in the analysis of the unexpected ductilization upon aging observed for some of the specimens: Two alloys reveal a surprising increase both in strength and in the total elongation after aging, namely, the lean maraging steel with 9 wt.% Mn (no or little retained austenite), Figs. 2b, 3b and the maraging TRIP steel with 12 wt.% Mn (containing retained austenite), Figs. 2c,3c,7.

For the 9 wt.% Mn maraging steel we observe that the UTS increases by 25-30% to more than 1 GPa and the TE by more than 150% (from 6% TE to more than 15% TE) due to the aging heat treatment (450°C for 48 hours). For the 12 wt.% Mn maraging TRIP alloy we observe after aging an increases in UTS to more than 1.3 GPa and of the TE from 16% to more than 21%. This increase in both mechanical properties (UTS, TE) represents a very unusual feature of these ultrahigh strength materials, Fig. 1. All other strengthening methods explored so far in the field of ultra high strength steels lead to a decrease in the ductility rather than to its enhancement.[34]

The difference in the retained austenite content between the two alloys (9 wt.% Mn, 12 wt.% Mn) means that the unexpected ductilization effect seems not to coincide with the occurrence of the TRIP effect, as one of the materials does by practical standards not seem to contain retained austenite (9 wt.% Mn). The increase in strength upon aging heat treatment is attributed to precipitation hardening. The TEM and APT results, Figs. 5 and 6, reveal that the steels contain nano-sized particles such as also found in conventional maraging steels[12-17] and that the particle dispersion is very high, even after the long aging treatment (450°C, 48h) used in this study (shown here exemplarily for the 12 wt.% Mn steel). The observation of a high maintained strength and high reluctance of the nanoscaled precipitates to coarsen was for conventional maraging steels reported before.[15] Using an Orowan line tension approximation for dislocation bow out between the nano-precipitates with a particle diameter of about 10 nm and an average particle spacing of about 100 nm suggests a potential increase in the yield strength upon aging of about 350 MPa. This increase in the initial yield strength roughly matches the change in properties observed. The effect of the high dispersion on further strain hardening during deformation, in particular on deformation homogeneity, is even stronger than its effect on strength as will be discussed in more detail below using a Kocks-Mecking analysis.

Figure 6: ATP measurements (IMAGO LEAP 3000x HR) conducted on the maraging
alloy with 9wt% Mn. (a) Atomic distribution of the Ni content (green). The shaded
zones indicate atomic Ni concentrations above 20 at.%. (b) Distribution of the Ni
(green), Al (yellow), and Ti (magenta) clusters together with Al atoms (yellow). The
shaded zones indicate an atomic Ni (green) concentration above 20 at.%, an atomic Al
(yellow) concentration above 10 at.%, and an atomic Ti (magenta) concentration above 5
at.%. (a) (b) Zoom Image
Figure 6: ATP measurements (IMAGO LEAP 3000x HR) conducted on the maraging
alloy with 9wt% Mn. (a) Atomic distribution of the Ni content (green). The shaded
zones indicate atomic Ni concentrations above 20 at.%. (b) Distribution of the Ni
(green), Al (yellow), and Ti (magenta) clusters together with Al atoms (yellow). The
shaded zones indicate an atomic Ni (green) concentration above 20 at.%, an atomic Al
(yellow) concentration above 10 at.%, and an atomic Ti (magenta) concentration above 5
at.%. (a) (b)
[less]

Besides this strong effect of the nano-precipitates formed during aging also the high retained dislocation content observed in the martensite matrix is important for the high strength. The TEM data revealed that the dislocation density in the martensite is very high, about 1015-1016 m-2. It is noteworthy that the dense dislocation arrangement prevailed even after the aging heat treatment. The dislocation density could also play an important role for the nucleation of the precipitates and their very high dispersion since we observed that many precipitates were located at dislocations. It is also important for the plastic properties that the (nearly) carbon-free martensite matrix is rather ductile. Conventional carbon-based martensitic steels typically reveal very poor ductility, Fig. 1.

However, irrespective of this plausible relationship between the nano-precipitates and the increase in flow stress, a central point of our observations is the simultaneous strong increase in the total elongation after aging, Figs. 2b,c. Two effects are conceivable to explain this phenomenon, namely, first, the kinetics associated with delayed austenitization and second, the kinetics of precipitation. Regarding austenitization it has been observed on conventional maraging steels that the heating rate and the holding times both influence the kinetics of re-transformation from martensite to austenite.[39] In other words it is well conceivable that very long heat treatment times (450°C, 48 h) entail delayed austenitization.
Dilatometry studies conducted at different heating rates confirm this phenomenon. Our data show the first deviation in length expansion from martensite upon re-heating at a temperature of about 550°C for the 12 wt.% Mn sample when using a quite rapid heating rate of 0.86K/s. The same experiment conducted at a heating rate of 0.086K/s revealed a lower transformation point from martensite into austenite around 500°C. Thermodynamic predictions also suggest a transformation around 480°C although some of the data underlying such simulations for Mn-containing Fe-based multi-component systems are not considered as very reliable.

Figure 7: High resolution EBSD maps taken on the 12 wt%Mn sample in the as-quenched plus aged state (a), and
the same microstructure at uniform elongation in the deformed flat tensile specimen (b). The top row shows the
phase content. The second row shows the texture in terms of the {hkl} Miller indices parallel to the normal
direction. The third row shows the texture in terms of the Miller indices parallel to the rolling direction.
The bottom row shows the average local misorientation (50 nm step size). Zoom Image
Figure 7: High resolution EBSD maps taken on the 12 wt%Mn sample in the as-quenched plus aged state (a), and
the same microstructure at uniform elongation in the deformed flat tensile specimen (b). The top row shows the
phase content. The second row shows the texture in terms of the {hkl} Miller indices parallel to the normal
direction. The third row shows the texture in terms of the Miller indices parallel to the rolling direction.
The bottom row shows the average local misorientation (50 nm step size).
[less]

In summary these considerations indicate that partial re-transformation into austenite might play a role for the ductilization at least for the specimen with 12 wt.% Mn. On the other hand the alloy with only 9 wt.% Mn shows the same ductilization effect although partial re-transformation at 450°C is thermodynamically not so likely for this alloy owing to it low austenite-stabilizing Mn content, Fig. 2b.

A somewhat more plausible explanation for the ductilization effect can, hence, be seen in the Orowan hardening mechanism. Fig. 8 shows the Kocks-Mecking analysis (strain hardening versus true stress curve) for the two grades with 9 wt.% Mn (no or little retained austenite) and 12 wt.% Mn (at least 15 vol.% retained austenite) together with the Considére line (strain hardening equals true stress). The data reveal in either case that the Orowan hardening effect is not active at the beginning of the tensile test but occurs as a second hardening plateau at larger stresses, Fig. 8. This plateau serves as a hardening reserve when the material starts to localize (Considére criterion).

The Kocks-Mecking curves for the as-quenched samples which were not heat treated and, hence, not aged, Fig. 8, do not show a second hardening plateau. A corresponding analysis on the conventional Ni-Co based maraging steel does not show this effect either before quenching. This confirms that the second hardening plateau can probably be interpreted in terms of Orowan hardening at least as the main mechanism. The reason that this effect does not occur at the beginning of straining is attributed to the fact that the inter-particle spacing is so small (~100 nm) that a higher stress level must be reached before Orowan loops can become active.

The Kocks-Mecking curves for the as-quenched samples which were not heat treated and, hence, not aged, Fig. 8, do not show a second hardening plateau. A corresponding analysis on the conventional Ni-Co based maraging steel does not show this effect either before quenching. This confirms that the second hardening plateau can probably be interpreted in terms of Orowan hardening at least as the main mechanism. The reason that this effect does not occur at the beginning of straining is attributed to the fact that the inter-particle spacing is so small (~100 nm) that a higher stress level must be reached before Orowan loops can become active.

Figure 8: Kocks–Mecking analysis (strain hardening vs. stress curve) of two alloys (9wt% Mn, 12 wt% Mn) which reveals the background for the increase in TE upon aging heat
treatment. Aging leads to a second hardening level at large stresses which does not appear in the as-quenched specimens. The intersections between the Conside´re lines (hardening
equals true stress) and the hardening curves indicate the onset of mechanical instability. Zoom Image
Figure 8: Kocks–Mecking analysis (strain hardening vs. stress curve) of two alloys (9wt% Mn, 12 wt% Mn) which reveals the background for the increase in TE upon aging heat
treatment. Aging leads to a second hardening level at large stresses which does not appear in the as-quenched specimens. The intersections between the Conside´re lines (hardening
equals true stress) and the hardening curves indicate the onset of mechanical instability.
[less]

Conclusions

We presented a concept for the development of ultrahigh strength and at the same time ductile martensitic and austenitic-martensitic steels that is based on the precipitation of nano-sized particles via aging heat treatment and, in some cases, on the TRIP effect. The approach, hence, combines the TRIP mechanism with a maraging treatment. The alloy systems presented are Fe-Mn steels with a low-carbon martensitic matrix and elements for the formation of nano-precipitates (Ni, Ti, Al, Mo). Particularly the 12 wt.% Mn maraging TRIP steel and the 9 wt.% Mn maraging steel revealed a significant increase both in strength and total elongation after the aging heat treatment. The unexpected increase in total elongation was mainly attributed to the formation of nanoscaled particles leading to an Orowan hardening mechanism and to a second strain hardening plateau at intermediate strains.

[1] J. Inoue, S. Nambu, Y. Ishimoto, T. Koseki, Scr. Mater. 2008, 59, 1055.
[2] S. Nambu, M. Michiuchi, Y. Ishimoto, K. Asakura, J. Inoue, T. Koseki, Scr. Mater. 2008, 59, 1055.
[3] J. R. Patel, M. Cohen, Acta Metall. 1953, 1, 531.
[4] H. K. D. H. Bhadeshia, D. V. Edmonds, Metall. Trans 1979, 10A, 895.
[5] G. B. Olson, M. Cohen, in: Mechanical Properties and Phase Transformations in Engineering Materials (Eds: S. C. Antolovich, R. O. Ritchie, W. W. Gerberich), New Orleans Met. Soc. AIME, 1986, pp. 367–390.
[6] M. Takahashi, H. K. D. H. Bhadeshia, Mater. Trans JIM 1991, 32, 689.
[7] P. J. Jacques, E. Girault, T. Catlin, N. Geerlofs, T. Kop, S. van der Zwaag, F. Delannay, Mater. Sci. Eng, A 1999, 273, 475.
[8] M. De Meyer, D. Vanderschueren, B. C. De Cooman, ISIJ Int. 1999, 39, 813.
[9] S. Traint, A. Pichler, K. Hauzenberger, P. Stiaszny, E. Werner, Steel Res. Int. 2002, 73, 259.
[10] P. J. Jacques, Curr. Opin. Solid State Mater. Sci. 2004, 8, 259.
[11] S. Zaefferer, J. Ohlert, W. Bleck, Acta Mater. 2004, 52, 2765.
[12] R. F. Decker, J. T. Eash, A. J. Goldman, Trans. ASM 1962, 55, 58.
[13] R. F. Decker, Source Book on Maraging Steels, ASM International, Metals Park, OH 1979.
[14] W. Sha, A. Cerezo, G. D. W. Smith, Metall. Trans. 1993, 24A, 1251.
[15] V. K. Vasudervan, S. J. Kim, C. M. Wayman, Metall. Trans. 1990, 21A, 2655.
[16] C. Servant, N. Bouzid, Acta Metall. 1988, 36, 2771.
[17] R. Tewari, S. Majumder, I. S. Batra, G. K. Dey, S. Banerjee, Acta Mater. 2000, 48, 1187.
[18] A. J. Goldman, J. Manenc, Trans. ASM 1965, 58, 645.
[19] R. K. Ray, A. K. Seal, Scr. Metall. 1976, 10, 971.
[20] M. Alinger, G. Odette, D. Hoelzer, Acta Mater. 2009, 57, 392.
[21] G. R. Odette, M. J. Alinger, B. D. Wirth, Ann. Rev. Mater. Res. 2008, 38, 471.
[22] D. R. Lesuer, C. K. Syn, O. D. Sherby, Mater. Trans. 2006, 47, 1508.
[23] S. Yamashita, S. Ohtsuka, N. Akasaka, S. Ukai, S. Ohnuki, Philos. Mag. Lett. 2004, 84, 525.
[24] M. Taneike, F. Abe, K. Sawada, Nature 2003, 424, 294.
[25] E. Courtois, T. Epicier, C. Scott, Micron 2006, 37, 492.
[26] A. Machova, Mater. Sci. Eng, A 2001, 319, 574.
[27] L. Storojeva, D. Ponge, R. Kaspar, D. Raabe, Acta Mater. 2004, 52, 2209.
[28] R. Song, D. Ponge, R. Kaspar, D. Raabe, Z. Metallk. 2004, 95, 513.
[29] L. Storojeva, D. Ponge, D. Raabe, R. Kaspar, Z. Metallk. 2004, 95, 1108.
[30] R. Song, D. Ponge, D. Raabe, R. Kaspar, Acta Mater. 2004, 53, 845.
[31] R. Song, D. Ponge, D. Raabe, Scr. Mater. 2005, 52, 1075.
[32] R. Song, D. Ponge, D. Raabe, ISIJ Int. 2005, 45, 1721.
[33] R. Song, D. Ponge, D. Raabe, Acta Mater. 2005, 53, 4881.
[34] R. Song, D. Ponge, D. Raabe, J. G. Speer, D. K. Matlock, Mater. Sci. Eng, A 2006, 441, 1.
[35] N. Tsuji, Y. Saito, H. Utsunomiya, S. Tanigawa, Scr. Mater. 1999, 40, 795.
[36] D. Raabe, D. Ponge, O. Dmitrieva, B. Sander, Scr. Mater. 2009, 60, 1141.
[37] K. Verbeken, N. Van Caenegem, D. Raabe, Micron 2009, 40, 151.
[38] G. P. Miller, W. I. Mitchell, J. Iron Steel Inst. 1965, 20, 899.
[39] R. Kapoor, I. S. Batra, Mater. Sci. Eng, A 2004, 371, 324.

 
Go to Editor View
loading content