- Email: [email protected]
Microstructure evolution and texture development in a friction stir-processed AISI D2 tool steel
Applied Surface Science 293 (2014) 151–159
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Microstructure evolution and texture development in a friction stir-processed AISI D2 tool steel N. Yasavol a , A. Abdollah-zadeh a,∗ , M.T. Vieira b , H.R. Jafarian c a b c
Department of Materials Engineering, Tarbiat Modares University, P.O. Box 14115-143, Tehran, Iran CEMUC, Department of Mechanical Engineering, University of Coimbra, P.O. Box 3030-788, Coimbra, Portugal School of Metallurgy and Materials Engineering, Iran University of Science and Technology (IUST), Tehran, Iran
a r t i c l e
i n f o
Article history: Received 9 September 2013 Received in revised form 11 December 2013 Accepted 21 December 2013 Available online 29 December 2013 Keywords: Tool steel EBSD Dynamic recrystallization Particle-stimulated nucleation Residual stress Texture
a b s t r a c t Crystallographic texture developments during friction stir processing (FSP) of AISI D2 tool were studied with respect to grain sizes in different tool rotation rates. Comparison of the grain sizes in various rotation rates confirmed that grain refinement occurred progressively in higher rotation rates by severe plastic deformation. It was found that the predominant mechanism during FSP should be dynamic recovery (DRV) happened concurrently with continuous dynamic recrystallization (CDRX) caused by particlestimulated nucleation (PSN). The developed shear texture relates to the ideal shear textures of D1 and D2 in bcc metals. The prevalence of highly dense arrangement of close-packed planes of bcc and the lowest Taylor factor showed the lowest compressive residual stress which is responsible for better mechanical properties compared with the grain-precipitate refinement. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Microstructure and texture evolution is one of the imperative key issues to have a better understanding of friction stir welding/processing (FSW/P). Knowing the potential advantages of these two technologies, numerous research efforts have recently been made to understand the details of structural progress during FSW/FSP; so far, such research attempts have been focused mainly on fcc metals such as aluminium alloys [1–3]. On the contrary, a reduced amount of attention has been given to bcc materials [4,5]. It has convincingly been established that the microstructure evolution during FSW/P is a very complex process driven by geometric effects of strain, grain subdivision and thermally activated high angle grain-boundary migration. The electron back-scattered diffraction (EBSD) technique has been considered as a powerful tool, developing better understanding of microstructural advancements and it also enhanced prediction behaviour of material after FSW/P [6]. Although many authors have claimed that FSW/P leads to formation of a refined, low-aspect-ratio grain structure including a significant fraction of high angle boundaries (HABs) [7,8]. However, other studies [9,10] have reported that considerable amount
∗ Corresponding author. Tel.: +98 21 88005040; fax: +98 21 88005040. E-mail address: [email protected] (A. Abdollah-zadeh). 0169-4332/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.apsusc.2013.12.122
of retained low angle boundaries (LABs) (up to 40%) exists after FSW/P of steels. It has also been revealed that the strain field encircling the rotation tool is predominantly simple shear in nature, so the texture progressed in the stir zone may be described in terms of ideal simple shear components, with the shear plane normal and shear direction aligned approximately perpendicular to and tangential with either the pin column surface or the flow lines in the stir zone, respectively [11,12]. Because of lower temperature experienced during FSW/P compared to conventional welding technique and surface treating, lower residual stresses remain in welded or processed materials [12]. However, there are different opinions about dominant source of the residual stress. For instance, previous study [13] reported that both friction heat and plastic deformation introduce residual stress during FSW of stainless steel. Thereafter, Woo et al. [14] proved that in FSP of 6061 aluminium alloy, the plastic deformation has rather effected than the frictional heating. On the contrary, Buffa et al. [15] proved that the effect of generated heat is more dominant. It calls for a comprehensive study of the effect of FSP rotation rate on residual stress of processed samples in order to find out the origin of residual stress and the consequent properties. Alternatively, Hatamleh et al. [16] have related these different residual stresses results to different material thicknesses, used parameters and the kind of clamping fixtures used to hold the plates in place during FSW/P. Considering these controversial claims about microstructure revolution comprising grain boundaries style, residual stress
152
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
Fig. 1. Schematic of the FSP setup. The rectangular shows the location of the microanalysis.
and obtained texture after FSW/P needs to be intensively investigated as to how these revolutions happen during vigorous plastic deformations at high temperatures. Although, previous researches claimed that grain refinement during FSW/P is responsible for mechanical properties improvement [17,18], our previous studies in nanohardness measurements [19] showed that the grain refinement does not accompany with improved mechanical properties. Consequently, it seems that not only microstructure refinement, but also texture plays a significant role on the mechanical properties. Furthermore, relatively little attention has been paid to how grain refinement occurs and what kinds of phenomena are dominant for the grain refinement during FSP. The present study focuses on the mechanisms of grain refinement during FSP of AISI D2 steel, and also, the related texture. For these purposes, the EBSD technique in conjunction with scanning electron microscopy equipped with a field emission gun (FESEM) was employed to provide in-depth insight into microstructural evolution. 2. Experimental procedure Sheets of as-annealed AISI D2 cold-worked die steel, with the chemical composition of 11.40Cr–1.49C–0.82Mo–0.79V– 0.40Si–0.35Mn–0.31Ni–bal Fe (in wt pct), were used. The top surfaces of the plates were mechanically ground using 80-grit emery cloth to remove oxide and contaminants, degreased with ethanol and then subjected to FSP setup. The FSP tool, made of WC-Co, had a columnar shape with a 16 mm diameter shoulder and no pin. Fig. 1 shows schematic image of FSP with the defined directions of normal direction (ND), transverse direction (TD) and rolling direction (RD) and the selected area of microanalysis studies. The FSP was accomplished in position control, using 3◦ tool tilt angle (˛) and 0.1 mm tool penetration into 3 mm thick AISI D2 plates. A constant traverse speed of 385 mm/min and four different tool rotation rates of 400, 500, 600 and 800 rpm were used as the processing parameters. Ar gas was used for surface shielding during FSP. As-processed specimens were sectioned transverse to the FSP direction and mechanically polished and etched with Nital solution. The microstructures of the specimens were characterized by optical microscope and scanning electron microscopy (SEM) using a field-emission type gun (Hitachi S-4300SE) equipped with electron backscatter diffraction (EBSD) system operated at 15 kV. The section of the specimen for microstructural observation was polished mechanically and then polished under the silicon colloidal solution. The obtained EBSD data were analysed by TSL-OIM analysis software ver. 5.3. All the microstructures of the FSPed specimens were observed from TD of the sheets. Using Cameca Camebax SX50 apparatus (Cameca, Gennevilliers, France), electron probe microanalysis (EPMA) was done to quantitatively measure the composition across the matrix and the carbide particles of FSPed samples. X-ray diffraction was used to measure the residual stress in samples using the multiple exposure technique. Residual stress measurements accomplished on FSPed samples with X-ray diffraction (XRD) with
Fig. 2. OM image of the BM.
Cr–K␣ radiation with a mean penetration depth of 5 m from surface. Since the primary and largest residual stress in FSW/P occurs in the longitudinal or weld-line direction [13,15], in this research, the residual stress was measured in longitudinal direction. The residual stress was calculated based on the general Hook’s law [20]. 3. Results and discussion An optical microscope image of the base material (BM), exhibiting the ferrite matrix and the large carbide particles, is illustrated in Fig. 2. The SEM images of stir zones (SZs) of the FSPed samples at different rotation rates of 400–600 rpm are shown in Fig. 3(a)–(c). They indicate that all regions consist of a microstructure having the ferrite matrix and fragmented carbides originated from the primary large carbides of the BM. In Fig. 3(d)–(f), the ND orientation colour maps of SZs with different tool rotation rates of 400–600 rpm obtained by EBSD measurements and analysis are shown. The colours in the orientation maps indicate the crystallographic directions of each point parallel to the transverse direction (TD) of the sheet (the normal direction of the observed plane), according to the stereographic triangle shown in the figure. It has been demonstrated that using tool with no pin during FSP, the shear stress generated by the forward motion of the tool works is concentrated near the upper surface of the SZ [21]. So, the shear plane normal (SPN) is always considered to be parallel to ND, while shear directions (SDs) are supposed to be tangential to SZ’s boundary in each point. The direction of the shear stress imposed on the SZ after FSP is representative by grain orientations which is much more extensive in the SZ of the FSPed sample at rotation rate of 500 rpm (Fig. 3(e)). In fact, in this condition, FSP arrangement resulted in the formation of high density close-packed planes (CPPs) [22] (shown in green colour) aligned almost 45◦ to TD in all over the SZ. Comparing the SZ of the FSPed sample at rotation rate of 500 rpm with the SZs of the lower and higher rotation rates of 400 and 600 rpm exhibits lower density of CPPs and higher density of {1 0 0} and {1 1 1} (Fig. 3(d)–(f)). Considering the SZ of 600 rpm FSPed sample, it also seems that lengthening of the grains has been promoted by effective severe plastic deformation and friction heat obtained at higher rotation rates. It can be implied that the grain lengthening is caused by geometric requirements of strain and grain boundary sliding which is happening during superplastic deformation [23]. In grain boundary maps of Fig. 3(g)–(i), the LABs and the HABs are depicted as blue and red lines, respectively. It is obvious that the microstructure of all FSPed samples consists of a mixture of the
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
153
Fig. 3. SEM images of the SZs of FSPed samples, shown in Fig. 1, at rotation rates of (a) 400, (b) 500, (c) 600, and (d) 800 rpm; IPF EBSD maps of the selected areas shown in the related SEM images of different rotation rates. The individual grains are coloured according to their crystallographic directions relative to WD; colour code triangle is shown in the bottom right corner. LABs and HABs are depicted as blue and red lines, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)
LABs and the HABs, as frequently observed after dynamic recrystallization (DRX) [24]. Furthermore, the LABs are distributed almost homogeneously within the grains and along the original grain boundaries, causing the formation of small equiaxed subgrains. As shown in Table 1, the fraction of HABs decreased from 81.1% to 75.9% and 78.0% for rotation rates of 400–500 and 600 rpm, accordingly. However, it dropped remarkably to 62.4% for rotation rate of 800 rpm, indicating that fraction of LABs increases by increasing the rotation rate. In the case of FSPed SZs of 400–600 rpm, the obtained microstructures are the same as true random grain assembly (80–90%) shown by previous reports [9,10]. It can be implied that the SZ microstructure of FSPed at 800 rpm with the highest amount of LABs represents appearance of dynamic recovery (DRV). These phenomena have also been reported in FSW of AA6082-T6 Table 1 Grain boundary fraction of FSPed AISI D2 at different rotation rates. Rotation rare (rpm)
Grain boundary fraction (2–15◦ )
Grain boundary fraction (15–180◦ )
400 500 600 800
0.189 0.241 0.220 0.376
0.811 0.759 0.780 0.624
and AA7108-T79 [25] and AA5754 [26]. On the other hand, the extreme alteration of the fraction of the HABs at higher rotation rate of 800 rpm correlated to enormous plastic deformation of FSPed samples [27]. In addition, SZ of FSPed sample at rotation rate of 800 rpm was difficult to characterize because the quality of EBSD pattern becomes severely degraded due to residual strains [28]. This is also a reflection of the steep strain and strain rate gradients in such locations, which is accompanied with increasing LABs population. To provide additional perception into microstructure development, misorientation distribution profiles were extracted from the EBSD measurements and are arranged as in Fig. 4. The misorientation distribution profiles of FSPed samples at rotation rates of 400–600 rpm are qualitatively similar to each other, and microstructure mostly consists of HABs having the average misorientation angle of 14◦ , 47◦ and 51◦ and an almost broad peak in high-angle ranges. This misorientation distribution profile of the FSPed AISI D2 is close to a random grain assembly predicted by Liu et al. [29]. The misorientation distribution profile of FSPed sample at rotation rate of 800 rpm shows that the average misorientation angle of 12◦ , 22–33◦ and also a more intensified broad peak at angle of 44 degrees. It seems that the subdivision process results from geometrically necessary dislocation boundaries
154
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
Fig. 6. SEM image of the SZ of FSPed samples at rotation rate of 500 rpm. Fig. 4. Boundary misorientation angle distribution of the SZs of FSPed samples at different rotation rates.
acquired by operative slip systems activation inside the elongated grains during FSP. Considering comprehensive dependence of grain boundaries development via the grain subdivision mechanism to texture development, it is believed that deviating rotation of subgrain orientations towards different stable texture components (defined by a given deformation mode) causes a continuous increase in average misorientation angle of LABs and thus to their gradual transformation into HABs. Therefore, development of the grain structure is a complex process involving the geometric effects of strain and local grain boundaries migration [30]. Fig. 5 shows distribution of grain size in the SZ of all FSPed samples. According to grain size distributions, the gradient in grain size between the BM and the SZ is much stronger than that obtained between the BM and the SZ of Al reported by Coelho et al. [31]. From this figure, it is evident that in SZs of FSPed samples at rotation rates of 400–600 rpm, there is a monotonic decreasing trend variation in grain size from 150 to 400 nm, whereas the SZ
Fig. 5. Grain size distribution of the SZs of FSPed samples at different rotation rates.
microstructure of FSPed sample at rotation rate of 800 rpm exhibits a microstructure having grain size of 100 nm. Our results are in disagreement with previous work done by Ueji et al. [10] and Hassan et al. [32], who reported that the overall grain size increases with increasing rotation rates. These conflicts may be attributed to the predominant deformation mechanism occurring during FSP of AISI D2 and its influence on grain growth. Since DRV is characteristic of metals with high stacking fault energy experiencing hot working such as Al alloys and ␣-Fe, it can be implied that the main occurrence mechanism during FSP of AISI D2 tool steel should be DRV [33]. As shown in Fig. 3(g)–(i) and Table 1, the concurrent existence of LABs and HABs and the elongated deformed grains confirm DRV mechanism during FSP of AISI D2. Beside this, it is well known that work hardening due to dislocation multiplication and recovery due to dislocation rearrangement are two competitive phenomena during hot working providing the steady state condition, dynamic equilibrium, reflecting equiaxed subgrain formation. Therefore, considering the highest amount of HABs and the equiaxed grain formation of Fig. 3(g) and Table 1, it can be perceived that the steady state condition was obtained in FSP at 400 rpm rotation rate. Whereas few amounts of HABs in the EBSD misorientation maps of Fig. 3(h), (i) and Table 1, it can be implied that during FSP conditions at rotation rates of 500–800 rpm, work hardening must be predominant phenomenon than work softening [22,34]. As shown in Fig. 6, the large non-deformable carbide particles (>1 m) are surrounded with the fine grains. These areas associating with a zone of lattice rotation, which is due to accumulation of the concentrated stress, are known as the particle-stimulated nucleation (PSN) and thereby CDRX mechanism. Sato et al. [35] have also reported the nucleation (CDRX mechanism) of the new grains during FSW of precipitation-hardened 6063 Al. It is well known that grain subdivision is essentially controlled by rotation rate of different parts of the same original grains towards different texture components [30]. Hence it is needed to study texture development in the FSPed samples. Fig. 7(a)–(d) shows {1 1 0} pole figures obtained from EBSD analysis of the FSPed specimens. In order to evaluate the preferred texture, the positions of ideal bcc shear texture components are elucidated in Fig. 7(e). In Fig. 7(a)–(d), the almost symmetric simple shear texture, which is characteristic of DRX attained from severe plastic deformation of FSW/P [36], is clearly apparent. This figure also reveals that the progressive variation of the texture intensity is accompanied with increasing the rotation rate of FSP in such a way that from rotation rate of 400–800 rpm goes along with monotonic increase
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
155
Fig. 7. (1 1 0) pole figures showing the bcc ferrite texture evolution as a function of FSP rotation rates of (a) 400, (b) 500, (c) 600, (d) 800 rpm and (e) the ideal bcc shear texture components. Note that each pole figure has its own scale bar and provides a qualitative observation of the texture evolution. In all pole figures the North Pole denotes the shear plane normal and east the shear plane direction as RD and TD of the SZs.
strain caused by the tool. It can be resulted that increasing the rotation rate of FSP, introducing higher strain and strain rate to the material, is associated with high number of independent slip systems activation, resulting in an increase of the texture intensity [36]. Furthermore, since the recovery processes involve a shortrange interaction between dislocations and subgrain boundaries, or between adjacent boundaries, they may lead to higher sliding on crystallographic planes and rotation of the lattice so that higher amount of the planes and directions are aligned towards the preferred directions (SDs) (preserving geometrical characteristics of plastic deformation) [36]. Therefore, a sharpening of deformation textures and a higher intensity of these deformation texture components appear, as given in Fig. 7(a)–(d), through increasing the rotation rate of FSP.
By increasing the FSP rotation rate, a clear systematic rotation of the texture can be observed about ND in such a way that in 400 rpm pole figures, the 1 1 0 poles are along with the RD (i.e. the distribution of the most of the (1 1 0) planes are parallel to RD). In the rotation rate of 500 rpm pole figures, the 1 1 0 poles are concentrated along with the TD (i.e. most of the (1 1 0) planes are perpendicular to RD). In the rotation rate of 600 rpm, the orientation 1 1 0 is the same as that of the 400 rpm but in upside down direction. The 1 1 0 poles of the rotation rate of 800 rpm are located by 45◦ in counter-close-wise direction with respect to RD as shown in Fig. 7(d). The systematic rotation of texture was previously shown across the advance through retreating sides of the FSW of 2519 Al alloy by Fonda and Bingert [37], and in FSW of AA6082-T651 by Ahmed et al. [38]. The dominant texture
156
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
Fig. 8. Orientation distribution function (ODF) of the SZs of the FSPed samples as a function of tool rotation rates. The ideal shear texture components of bcc metals and the custom prominent texture fibres.
components of all FSPed samples as indicated by the related pole figures in Fig. 7(a)–(d) mostly include D1, D2 which come from usual shear texture of rolling and recrystallization of bcc metals from Table 2 [39,40]. In this regard, Liu et al. [29] showed that the shear component is typically formed with inhomogeneous shear deformation due to the FSP. With the intention of more precisely assessing the texture and FSP rotation rates on dominance of texture evolution, the orientation distribution function (ODF) was calculated from the three experimentally measured, pole figures for each SZ. The ODFs of all
Table 2 Main ideal orientations in simple shear deformation of bcc materials [49]. Notation
{h k l} [u v w]
D1 D2 E1 E2 J1 J2 F
¯ [1 1 1] (1 1 2) (1¯ 1¯ 2) [1 1 1] ¯ [1 1 1] (0 1 1) (0 1¯ 1) [1 1 1] (0 1¯ 1) [2¯ 1 1] (1 1¯ 0) [1¯ 1¯ 1] (1 1 0) [0 0 1]
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
157
Table 3 EPMA chemical composition of AISI D2 elements in BM and FSPed sample. Element process BM FSP 600 rpm
Fig. 9. Orientation density along the different fibres versus ˚ and ϕ1 Euler angles, as a function of FSPed samples rotation rates.
FSPed samples are shown in Fig. 8 and the texture components and their intensities can be identified. In ODFs of all FSPed samples, no continuous bands of orientations extend in the 2 direction. Fig. 9(a)–(c) shows the prevalence fibres variations consisting of ␣, ␥ and which are appearing during FSP with different rotation rates. According to Fig. 9(a), the intensity of ␣-fibre (1 1 0//RD) is decreased by increasing the FSP rotation rates from 400 to 800 rpm. The homogeneous orientation density variation along the ␥-fibre ((1 1 1)//ND) in FSPed samples of 400–600 rpm and high increment of this fibre in 800 rpm represents the so-called pancake formation model of grains. In addition, the -fibre (0 1 1//TD) is decreased by rotation rates. The 400 and 500 rpm, because of the higher ␣-fibre compared to those of 600 and 800 rpm, show more random texture. The 800 rpm with the highest ␥-fibre (including {1 1 1} 1 2 3 component) growth by consuming the ␣-fibres with respect to those
V
Cr
0.74
7.95 8.45
Mn
Fe
Mo
0.27
81.20 86.66
0.61
of 400 and 500 rpm and therefore processes high stored energy components [41]. Considering the above results, it can be also implied that, at lower rotation rate of 400 rpm with the lower strain/strain rate, the lower slip system activated just provides the newly fine equiaxed grains by CDRX. At rotation rate of 500 rpm, according to crystal plasticity theory [42], the enough strain/strain rate or strain-induced progressive rotation of subgrain boundaries provides suitable slip system to be activated. According to Taylor’s original model defined elsewhere [37], it can be realized that if the dominant shear texture components satisfies the higher critical resolved shear stress, to minimize the total internal strain while maintaining compatible strain between grains in a polycrystalline structure, the lowest Taylor factor (M) will be obtained. Indeed, the different formed textures during various FSP rotation rates caused different Taylor factor. In this regard, in 500 rpm, the ␣-fibre decreased or the {1 1 1} 1 1 0 developed significantly and resulted in continuous lattice alignment in such a way that the high CPP/Ds to PSN and consistent with SD of the ideal simple shear reference frame. By increasing the rotation rate from 500 to 800 rpm, the Taylor factor increases. These results are not accompanied with the grain refinement trend obtained for different rotation rates of 400–800 rpm, while they are directly accompanied with the nanohardness measurements of 683, 1207, 552 and 583 HV for rotation rates of 400, 500, 600, and 800 rpm, accordingly. The predominant effect of texture with respect to microstructure refinement proved by Hall–Pech equation was previously reported by Wang et al. [43] in FSP of Mg–Al–Zn alloy. This also confirms the results obtained from IPFs of Fig. 3(d)–(f). The perpendicular oriented {1 1 0} planes to RD in 500 rpm FSPed samples indicate that in this condition, the CPPs are extensively affected by the applied axial pressure of the shoulder and backing plate [44] and consequently was higher than that of 400 and 600 rpm. On the other hand, it can be implied that at 500 rpm, the CPPs with the highest Young’s modulus is another responsible factor for better mechanical properties of this sample compared with the other FSPed samples [45]. As shown by EBSD results (Fig. 3(d)–(f)), the increment maintenance of the shear-type deformation texture with rotation rates confirms that DRV is the predominant mechanism of FSP. Alternatively, the EPMA results of Table 3, indicating the increment presence of the interstitials in ferrite compared with that of the BM and providing a significant drag on dislocations, results in positive strain rate sensitivity. In this regard, increasing the tool rotation rate provides more interstitial elements diffusion as well as a much stronger texture at rotation rate of 800 rpm compared to those obtained at the rotation rates of 400–600 rpm. Previous works [46] also showed that FSW/P microstructural trend seems to be like control rolled processing in which interstitial elements have a strong effect on texture formation during warm rolling. In conclusion, the positive strain rate sensitivity leads to more heterogeneous deformation and results in a sharpened texture which was shown in Fig. 7(a)–(d). Furthermore, when slip contributions to strain are limited by restricted numbers of slip systems or by solute drag effects, and boundary migration is similarly retarded by solute, lattice rotation may develop progressively at grain boundaries as dislocations accumulate during straining. This may lead to continuous dynamic recrystallization (CDRX) or extended recovery during FSP of AISI D2.
158
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
particles is proposed as another mechanism during FSP. The EBSD’s analysis and the texture studies of samples indicate that the dominant source of the residual stress is plastic deformation. In 500 rpm with high amount of close-packed planes show the most plastic deformation and compressive residual stress. This sample with the {1 1 1} 1 1 0, {1 1 2} 1 1 1 and {1 1 0} 0 0 1 shear components and the lowest Taylor factor shows the highest nanohardness measurements. Therefore, the formed textures play more predominant effect on mechanical properties than grain refinement.
References
Fig. 10. Longitudinal and shear residual stresses along the processed direction at the surfaces of BM and FSPed samples at different rotation rates.
The longitudinal residual stresses of FSPed SZs shown in Fig. 10 lie in a range of 2–12% of 0.2% offset yield strength of the parent metal (illustrated in Guu’s work [47]). These values are somewhat smaller than those reported in previous literatures [20,48] of (20–70%), which were typically close to that of the dynamically recrystallized SZs. This inconsistency can be ascribed to the residual stress resource. As shown in Fig. 3(d)–(f), the elongated appearance of the new grains confirms that grain refining of the FSPed samples is more influenced by severe plastic deformation compared with the concurrent produced heat. In fact, the lower heat input or predominant stress state lead to such small residual stress maintenance in all FSPed samples. From 400 to 500 rpm rotation rate, the longitudinal residual stress has altered from tensile to compressive. However, from 500 to 600 and 800 rpm, it changed to higher level of tensile residual stresses. This is in good agreement with those obtained by Lombard et al. [48] in FSW of AA5083 and H321 alloy. The sharp reduction of tensile residual stress can be related to concurrent effect of the severe plastic deformation and generated heat of FSP at 500 rpm rotation rate respect to that of 400 rpm. Indeed, in FSPed region of FSPed sample at rotation rate of 500 rpm as previously indicated by EBSDs’ and texture studies, the highly dense CPPs in accompany with the usual shear slip system of bcc metals reduced the residual stress of this FSPed sample. This residual stress decrease can also be ascribed to this situation which satisfies the Taylor’s original model. In contrast, it is possible to conclude that in FSPed samples at rotation rates of 600 and 800 rpm, the higher plastic deformation existence (with higher LABs/dislocation density) led to increase of the longitudinal residual stresses to higher level. Continuous increase of the shear stress of Fig. 10 infers the ascending trend of severe plastic deformation with respect to FSP rotation rates. 4. Conclusions In the present study, crystallographic texture development and the mechanisms happening during FSP of AISI D2 steel have been studied. The EBSD analysis shows that DRV and CDRX are the predominant mechanisms of grain refinement. Decreasing the HABs with increasing the rotation rates implies that work hardening or DRV must be the predominant phenomenon than work softening. This can be due to (I) an increased amount of solute carbon in ferrite retarding the formation and migration of HABs, and (II) the increased trend of residual stress by rotation rates during FSP which indicate the increase trend of dislocation densities. In addition to DRV, CDRX caused by PSN of large non-shearable carbide
[1] W.A. Baeslack, K.V. Jata, T.J. Lienert, Structure, properties and fracture of friction stir welds in a high-temperature Al–8.5Fe–1.3V–1.7Si alloy (AA-8009), J. Mater. Sci. 41 (2006) 2939–2951. [2] V. Balasubramanian, Relationship between base metal properties and friction stir welding process parameters, Mater. Sci. Eng. A 480 (2008) 397–403. [3] M. Dumont, A. Steuwer, A. Deschamps, M. Peel, P. Withers, Microstructure mapping in friction stir welds of 7449 aluminium alloy using SAXS, Acta Mater. 54 (2006) 4793–4801. [4] A.K. Lakshminarayanan, V. Balasubramanian, An assessment of microstructure, hardness, tensile and impact strength of friction stir welded ferritic stainless steel joints, Mater. Des. 31 (2010) 4592–4600. [5] S. Noh, R. Kasada, A. Kimura, S.H.C. Park, S. Hirano, Microstructure and mechanical properties of friction stir processed ODS ferritic steels, J. Nucl. Mater. 417 (1) (2011) 245–248. [6] P.J. Hurley, F.J. Humphreys, The application of EBSD to the study of substructural development in a cold rolled single-phase aluminium alloy, Acta Mater. 51 (2003) 1087–1102. [7] R.W. Fonda, J.F. Bingert, Precipitation and grain refinement in a 2195 Al friction stir weld, Metall. Mater. Trans. 37 (2006) 3593–3594. [8] H. Fujii, L. Cui, M. Maeda, K. Nogi, Effect of tool shape on mechanical properties and microstructure of friction stir welded aluminum alloys, Mater. Sci. Eng. A 419 (2006) 25–31. [9] H. Fujii, L. Cui, N. Tsuji, M. Maeda, K. Nakata, K. Nogi, Friction stir welding of carbon steels, Mater. Sci. Eng. A 429 (2006) 50–57. [10] R. Ueji, H. Fujii, L. Cui, A. Nishioka, K. Kunishige, K. Nogi, Friction stir welding of ultrafine grained plain low-carbon steel formed by the martensite process, Mater. Sci. Eng. A 423 (2006) 324–330. [11] S.H.S. Park, Y.S. Sato, H. Kokawa, K. Okamoto, S. Hirano, M. Inagaki, Rapid formation of the sigma phase in 304 stainless steel during friction stir welding, Scr. Mater. 49 (2003) 1175–1180. [12] A. Reynolds, Structure, properties, and residual stress of 304L stainless steel friction stir welds, Scr. Mater. 48 (2003) 1289–1294. [13] X.K. Zhu, Y.J. Chao, Numerical simulation of transient temperature and residual stresses in friction stir welding of 304L stainless steel, J. Mater. Process. Technol. 146 (2004) 263–272. [14] W. Woo, H. Choo, D. Brown, S. Vogel, P. Liaw, Z. Feng, Texture analysis of a friction stir processed 6061-T6 aluminum alloy using neutron diffraction, Acta Mater. 54 (2006) 3871–3882. [15] G. Buffa, L. Fratini, S. Pasta, R. Shivpuri, On the thermo-mechanical loads and the resultant residual stresses in friction stir processing operations, CIRP Ann. – Manuf. Technol. 57 (2008) 287–290. [16] O. Hatamleh, I.V. Rivero, A. Maredia, Residual stresses in friction-stir-welded 2195 and 7075 aluminum alloys, Metall. Mater. Trans. 39 (2008) 2867–2874. [17] C.M. Hu, C.M. Lai, P.W. Kao, N.J. Ho, J.C. Huang, Solute-enhanced tensile ductility of ultrafine-grained Al–Zn alloy fabricated by friction stir processing, Scr. Mater. 60 (2009) 639–642. [18] L. Johannes, R. Mishra, Multiple passes of friction stir processing for the creation of superplastic 7075 aluminum, Mater. Sci. Eng. A 464 (2007) 255–260. [19] N. Yasavol, A. Abdollah-zadeh, D.M. Rodrigues, Microstructure and mechanical properties of friction stir processed AISI D2 tool steel, in: Second International Conference on Friction Stir Welding and Processing, 26–27th January, 2012. [20] W. Woo, Z. Feng, X.-L. Wang, S.A. David, Neutron diffraction measurements of residual stresses in friction stir welding: a review, Sci. Technol. Weld. Join. (2004) 23–32. [21] Y. Sato, H. Kokawa, K. Ikeda, M. Enomoto, S. Jogan, T. Hashimato, Microtexture in the friction-stir weld of an aluminum alloy, Metall. Mater. Trans. 32 (2008) 941–947. [22] K. Dehghani, A. Chabok, Dependence of Zener parameter on the nanograins formed during friction stir processing of interstitial free steels, Mater. Sci. Eng. A 528 (2011) 4325–4330. [23] W. Xu, J. Liu, G. Luan, C. Dong, Temperature evolution, microstructure and mechanical properties of friction stir welded thick 2219-O aluminum alloy joints, Mater. Des. 30 (2008) 1886–1893. [24] R.S. Mishra, M.W. Mahoney, Friction Stir Welding and Processing, ASM International, United States of America, 2007. [25] Ø. Frigaard, Ø. Grong, T. Mildingo, A process model for friction stir welding of age hardening aluminum alloys, Metall. Mater. Trans. 32A (2001) 1189–1200. [26] A. Gerlich, M. Yamamoto, T.H. North, Strain rates and grain growth in Al 5754 and Al 6061 friction stir spot welds, Metall. Mater. Trans. 38 (2007) 1291–1302.
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159 [27] A.P. Gerlich, T. Shibayanagi, Grain boundary sliding during friction stir spot welding of an aluminum alloy, Scr. Mater. 60 (2009) 236–239. [28] F.J. Humphreys, Characterisation of fine-scale microstructures by electron backscatter diffraction (EBSD), Scr. Mater. 51 (2004) 771–776. [29] F.C. Liu, Z.Y. Ma, L.Q. Chen, Low-temperature superplasticity of Al–Mg–Sc alloy produced by friction stir processing, Scr. Mater. 60 (2009) 968–971. [30] S. Mironov, Y.S. Sato, H. Kokawa, Microstructural evolution during friction stirprocessing of pure iron, Acta Mater. 56 (2008) 2602–2614. [31] R.S. Coelho, A. Kostka, S. Sheikhi, J. dos Santos, A.R. Pyzalla, Microstructure and mechanical properties of an AA6181-T4 auminium alloy to HC340LA high strength steel friction stir overlap weld, Adv. Eng. Mater. 10 (2008) 961–972. [32] K.A.A. Hassan, A.F. Norman, D.A. Price, P.B. Prangnell, Stability of nugget zone grain structures in high strength Al alloy friction stir welds during solution treatment, Acta Mater. 51 (2003) 1923–1936. [33] T.J. Lienert, W.L. Stellwag JR., B.B. Grimmett, R.W. Warke, Friction stir welding studies on mild steels, Weld. J. (2003) 1–9. [34] Y. Sato, H. Yamanoi, H. Kokawa, T. Furuhara, Microstructural evolution of ultrahigh carbon steel during friction stir welding, Scr. Mater. 57 (2007) 557–560. [35] Y. Sato, M. Urata, H. Kokawa, Parameters controlling microstructure and hardness during friction-stir welding of precipitation-hardenable aluminum alloy 6063, Mater. Sci. Eng. A 33 (2002) 625–635. [36] U.F. Kocks, C.N. Tomé, H.-R. Wenk, Texture and Anisotropy Preferred Orientations in Polycrystals and their Effect on Materials Properties, Cambridge University Press, United Kingdom, 1998. [37] R.W. Fonda, J.F. Bingert, Microstructural evolution in the heat-affected zone of a friction stir weld, Metall. Mater. Trans. 35 (2004) 1487–1499. [38] M.M.Z. Ahmed, B.P. Wynne, W.M. Rainforth, P.L. Threadgill, Quantifying crystallographic texture in the probe-dominated region of thick-section friction-stir-welded aluminium, Scr. Mater. 59 (2008) 507–510.
159
[39] I.L. Dillamore, W.T. Roberts, Rolling textures in FCC and BCC metals, Acta Metall. 12 (1964) 282–294. [40] M. Hoscher, D. Raabe, K. Lucke, Rolling and recrystallization textures of bcc steels, Steel Res. 62 (1991) 567–575. [41] B. Verlinden, J. Driver, I. Samajdar, R. Doherty, in: R.W. Cahn (Ed.), Thermo-Mechanical Processing of Metallic Materials, Elsevier, Amsterdam, 2007. [42] B. Heinz, B. Skrotzki, Characterization of a friction-stir-welded aluminum alloy 6013, Metall. Mater. Trans. 33 (2002) 497–504. [43] Y.N. Wang, C.I. Chang, C.J. Lee, H.K. Lin, J.C. Huang, Texture and weak grain size dependence in friction stir processed Mg–Al–Zn alloy, Scr. Mater. 55 (2006) 637–640. [44] S. Park, Y. Sato, H. Kokawa, Basal plane texture and flow pattern in friction stir weld of a magnesium alloy, Mater. Sci. Eng. A 34 (2003) 987–997. [45] W.D.J. Callister, Materials Science and Engineering: An Introduction, John Wiley and Sons, New York, 1940. [46] R. Nandan, T. DebRoy, H.K.D.H. Bhadeshia, Recent advances in friction-stir welding – process, weldment structure and properties, Prog. Mater. Sci. 53 (2008) 980–1023. [47] Y. Guu, AFM surface imaging of AISI D2 tool steel machined by the EDM process, Appl. Surf. Sci. 242 (2005) 245–250. [48] H. Lombard, D.G. Hattingh, A. Steuwer, M.N. James, Effect of process parameters on the residual stresses in AA5083-H321 friction stir welds, Mater. Sci. Eng. A 501 (2009) 119–124. [49] S. Lia, J.I. Beyerleinb, M.A.M. Bourkea, Texture formation during equal channel angular extrusion of fcc and bcc materials: comparison with simple shear, Mater. Sci. Eng. A 39 (2005) 66–77.
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Microstructure evolution and texture development in a friction stir-processed AISI D2 tool steel N. Yasavol a , A. Abdollah-zadeh a,∗ , M.T. Vieira b , H.R. Jafarian c a b c
Department of Materials Engineering, Tarbiat Modares University, P.O. Box 14115-143, Tehran, Iran CEMUC, Department of Mechanical Engineering, University of Coimbra, P.O. Box 3030-788, Coimbra, Portugal School of Metallurgy and Materials Engineering, Iran University of Science and Technology (IUST), Tehran, Iran
a r t i c l e
i n f o
Article history: Received 9 September 2013 Received in revised form 11 December 2013 Accepted 21 December 2013 Available online 29 December 2013 Keywords: Tool steel EBSD Dynamic recrystallization Particle-stimulated nucleation Residual stress Texture
a b s t r a c t Crystallographic texture developments during friction stir processing (FSP) of AISI D2 tool were studied with respect to grain sizes in different tool rotation rates. Comparison of the grain sizes in various rotation rates confirmed that grain refinement occurred progressively in higher rotation rates by severe plastic deformation. It was found that the predominant mechanism during FSP should be dynamic recovery (DRV) happened concurrently with continuous dynamic recrystallization (CDRX) caused by particlestimulated nucleation (PSN). The developed shear texture relates to the ideal shear textures of D1 and D2 in bcc metals. The prevalence of highly dense arrangement of close-packed planes of bcc and the lowest Taylor factor showed the lowest compressive residual stress which is responsible for better mechanical properties compared with the grain-precipitate refinement. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Microstructure and texture evolution is one of the imperative key issues to have a better understanding of friction stir welding/processing (FSW/P). Knowing the potential advantages of these two technologies, numerous research efforts have recently been made to understand the details of structural progress during FSW/FSP; so far, such research attempts have been focused mainly on fcc metals such as aluminium alloys [1–3]. On the contrary, a reduced amount of attention has been given to bcc materials [4,5]. It has convincingly been established that the microstructure evolution during FSW/P is a very complex process driven by geometric effects of strain, grain subdivision and thermally activated high angle grain-boundary migration. The electron back-scattered diffraction (EBSD) technique has been considered as a powerful tool, developing better understanding of microstructural advancements and it also enhanced prediction behaviour of material after FSW/P [6]. Although many authors have claimed that FSW/P leads to formation of a refined, low-aspect-ratio grain structure including a significant fraction of high angle boundaries (HABs) [7,8]. However, other studies [9,10] have reported that considerable amount
∗ Corresponding author. Tel.: +98 21 88005040; fax: +98 21 88005040. E-mail address: [email protected] (A. Abdollah-zadeh). 0169-4332/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.apsusc.2013.12.122
of retained low angle boundaries (LABs) (up to 40%) exists after FSW/P of steels. It has also been revealed that the strain field encircling the rotation tool is predominantly simple shear in nature, so the texture progressed in the stir zone may be described in terms of ideal simple shear components, with the shear plane normal and shear direction aligned approximately perpendicular to and tangential with either the pin column surface or the flow lines in the stir zone, respectively [11,12]. Because of lower temperature experienced during FSW/P compared to conventional welding technique and surface treating, lower residual stresses remain in welded or processed materials [12]. However, there are different opinions about dominant source of the residual stress. For instance, previous study [13] reported that both friction heat and plastic deformation introduce residual stress during FSW of stainless steel. Thereafter, Woo et al. [14] proved that in FSP of 6061 aluminium alloy, the plastic deformation has rather effected than the frictional heating. On the contrary, Buffa et al. [15] proved that the effect of generated heat is more dominant. It calls for a comprehensive study of the effect of FSP rotation rate on residual stress of processed samples in order to find out the origin of residual stress and the consequent properties. Alternatively, Hatamleh et al. [16] have related these different residual stresses results to different material thicknesses, used parameters and the kind of clamping fixtures used to hold the plates in place during FSW/P. Considering these controversial claims about microstructure revolution comprising grain boundaries style, residual stress
152
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
Fig. 1. Schematic of the FSP setup. The rectangular shows the location of the microanalysis.
and obtained texture after FSW/P needs to be intensively investigated as to how these revolutions happen during vigorous plastic deformations at high temperatures. Although, previous researches claimed that grain refinement during FSW/P is responsible for mechanical properties improvement [17,18], our previous studies in nanohardness measurements [19] showed that the grain refinement does not accompany with improved mechanical properties. Consequently, it seems that not only microstructure refinement, but also texture plays a significant role on the mechanical properties. Furthermore, relatively little attention has been paid to how grain refinement occurs and what kinds of phenomena are dominant for the grain refinement during FSP. The present study focuses on the mechanisms of grain refinement during FSP of AISI D2 steel, and also, the related texture. For these purposes, the EBSD technique in conjunction with scanning electron microscopy equipped with a field emission gun (FESEM) was employed to provide in-depth insight into microstructural evolution. 2. Experimental procedure Sheets of as-annealed AISI D2 cold-worked die steel, with the chemical composition of 11.40Cr–1.49C–0.82Mo–0.79V– 0.40Si–0.35Mn–0.31Ni–bal Fe (in wt pct), were used. The top surfaces of the plates were mechanically ground using 80-grit emery cloth to remove oxide and contaminants, degreased with ethanol and then subjected to FSP setup. The FSP tool, made of WC-Co, had a columnar shape with a 16 mm diameter shoulder and no pin. Fig. 1 shows schematic image of FSP with the defined directions of normal direction (ND), transverse direction (TD) and rolling direction (RD) and the selected area of microanalysis studies. The FSP was accomplished in position control, using 3◦ tool tilt angle (˛) and 0.1 mm tool penetration into 3 mm thick AISI D2 plates. A constant traverse speed of 385 mm/min and four different tool rotation rates of 400, 500, 600 and 800 rpm were used as the processing parameters. Ar gas was used for surface shielding during FSP. As-processed specimens were sectioned transverse to the FSP direction and mechanically polished and etched with Nital solution. The microstructures of the specimens were characterized by optical microscope and scanning electron microscopy (SEM) using a field-emission type gun (Hitachi S-4300SE) equipped with electron backscatter diffraction (EBSD) system operated at 15 kV. The section of the specimen for microstructural observation was polished mechanically and then polished under the silicon colloidal solution. The obtained EBSD data were analysed by TSL-OIM analysis software ver. 5.3. All the microstructures of the FSPed specimens were observed from TD of the sheets. Using Cameca Camebax SX50 apparatus (Cameca, Gennevilliers, France), electron probe microanalysis (EPMA) was done to quantitatively measure the composition across the matrix and the carbide particles of FSPed samples. X-ray diffraction was used to measure the residual stress in samples using the multiple exposure technique. Residual stress measurements accomplished on FSPed samples with X-ray diffraction (XRD) with
Fig. 2. OM image of the BM.
Cr–K␣ radiation with a mean penetration depth of 5 m from surface. Since the primary and largest residual stress in FSW/P occurs in the longitudinal or weld-line direction [13,15], in this research, the residual stress was measured in longitudinal direction. The residual stress was calculated based on the general Hook’s law [20]. 3. Results and discussion An optical microscope image of the base material (BM), exhibiting the ferrite matrix and the large carbide particles, is illustrated in Fig. 2. The SEM images of stir zones (SZs) of the FSPed samples at different rotation rates of 400–600 rpm are shown in Fig. 3(a)–(c). They indicate that all regions consist of a microstructure having the ferrite matrix and fragmented carbides originated from the primary large carbides of the BM. In Fig. 3(d)–(f), the ND orientation colour maps of SZs with different tool rotation rates of 400–600 rpm obtained by EBSD measurements and analysis are shown. The colours in the orientation maps indicate the crystallographic directions of each point parallel to the transverse direction (TD) of the sheet (the normal direction of the observed plane), according to the stereographic triangle shown in the figure. It has been demonstrated that using tool with no pin during FSP, the shear stress generated by the forward motion of the tool works is concentrated near the upper surface of the SZ [21]. So, the shear plane normal (SPN) is always considered to be parallel to ND, while shear directions (SDs) are supposed to be tangential to SZ’s boundary in each point. The direction of the shear stress imposed on the SZ after FSP is representative by grain orientations which is much more extensive in the SZ of the FSPed sample at rotation rate of 500 rpm (Fig. 3(e)). In fact, in this condition, FSP arrangement resulted in the formation of high density close-packed planes (CPPs) [22] (shown in green colour) aligned almost 45◦ to TD in all over the SZ. Comparing the SZ of the FSPed sample at rotation rate of 500 rpm with the SZs of the lower and higher rotation rates of 400 and 600 rpm exhibits lower density of CPPs and higher density of {1 0 0} and {1 1 1} (Fig. 3(d)–(f)). Considering the SZ of 600 rpm FSPed sample, it also seems that lengthening of the grains has been promoted by effective severe plastic deformation and friction heat obtained at higher rotation rates. It can be implied that the grain lengthening is caused by geometric requirements of strain and grain boundary sliding which is happening during superplastic deformation [23]. In grain boundary maps of Fig. 3(g)–(i), the LABs and the HABs are depicted as blue and red lines, respectively. It is obvious that the microstructure of all FSPed samples consists of a mixture of the
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
153
Fig. 3. SEM images of the SZs of FSPed samples, shown in Fig. 1, at rotation rates of (a) 400, (b) 500, (c) 600, and (d) 800 rpm; IPF EBSD maps of the selected areas shown in the related SEM images of different rotation rates. The individual grains are coloured according to their crystallographic directions relative to WD; colour code triangle is shown in the bottom right corner. LABs and HABs are depicted as blue and red lines, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)
LABs and the HABs, as frequently observed after dynamic recrystallization (DRX) [24]. Furthermore, the LABs are distributed almost homogeneously within the grains and along the original grain boundaries, causing the formation of small equiaxed subgrains. As shown in Table 1, the fraction of HABs decreased from 81.1% to 75.9% and 78.0% for rotation rates of 400–500 and 600 rpm, accordingly. However, it dropped remarkably to 62.4% for rotation rate of 800 rpm, indicating that fraction of LABs increases by increasing the rotation rate. In the case of FSPed SZs of 400–600 rpm, the obtained microstructures are the same as true random grain assembly (80–90%) shown by previous reports [9,10]. It can be implied that the SZ microstructure of FSPed at 800 rpm with the highest amount of LABs represents appearance of dynamic recovery (DRV). These phenomena have also been reported in FSW of AA6082-T6 Table 1 Grain boundary fraction of FSPed AISI D2 at different rotation rates. Rotation rare (rpm)
Grain boundary fraction (2–15◦ )
Grain boundary fraction (15–180◦ )
400 500 600 800
0.189 0.241 0.220 0.376
0.811 0.759 0.780 0.624
and AA7108-T79 [25] and AA5754 [26]. On the other hand, the extreme alteration of the fraction of the HABs at higher rotation rate of 800 rpm correlated to enormous plastic deformation of FSPed samples [27]. In addition, SZ of FSPed sample at rotation rate of 800 rpm was difficult to characterize because the quality of EBSD pattern becomes severely degraded due to residual strains [28]. This is also a reflection of the steep strain and strain rate gradients in such locations, which is accompanied with increasing LABs population. To provide additional perception into microstructure development, misorientation distribution profiles were extracted from the EBSD measurements and are arranged as in Fig. 4. The misorientation distribution profiles of FSPed samples at rotation rates of 400–600 rpm are qualitatively similar to each other, and microstructure mostly consists of HABs having the average misorientation angle of 14◦ , 47◦ and 51◦ and an almost broad peak in high-angle ranges. This misorientation distribution profile of the FSPed AISI D2 is close to a random grain assembly predicted by Liu et al. [29]. The misorientation distribution profile of FSPed sample at rotation rate of 800 rpm shows that the average misorientation angle of 12◦ , 22–33◦ and also a more intensified broad peak at angle of 44 degrees. It seems that the subdivision process results from geometrically necessary dislocation boundaries
154
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
Fig. 6. SEM image of the SZ of FSPed samples at rotation rate of 500 rpm. Fig. 4. Boundary misorientation angle distribution of the SZs of FSPed samples at different rotation rates.
acquired by operative slip systems activation inside the elongated grains during FSP. Considering comprehensive dependence of grain boundaries development via the grain subdivision mechanism to texture development, it is believed that deviating rotation of subgrain orientations towards different stable texture components (defined by a given deformation mode) causes a continuous increase in average misorientation angle of LABs and thus to their gradual transformation into HABs. Therefore, development of the grain structure is a complex process involving the geometric effects of strain and local grain boundaries migration [30]. Fig. 5 shows distribution of grain size in the SZ of all FSPed samples. According to grain size distributions, the gradient in grain size between the BM and the SZ is much stronger than that obtained between the BM and the SZ of Al reported by Coelho et al. [31]. From this figure, it is evident that in SZs of FSPed samples at rotation rates of 400–600 rpm, there is a monotonic decreasing trend variation in grain size from 150 to 400 nm, whereas the SZ
Fig. 5. Grain size distribution of the SZs of FSPed samples at different rotation rates.
microstructure of FSPed sample at rotation rate of 800 rpm exhibits a microstructure having grain size of 100 nm. Our results are in disagreement with previous work done by Ueji et al. [10] and Hassan et al. [32], who reported that the overall grain size increases with increasing rotation rates. These conflicts may be attributed to the predominant deformation mechanism occurring during FSP of AISI D2 and its influence on grain growth. Since DRV is characteristic of metals with high stacking fault energy experiencing hot working such as Al alloys and ␣-Fe, it can be implied that the main occurrence mechanism during FSP of AISI D2 tool steel should be DRV [33]. As shown in Fig. 3(g)–(i) and Table 1, the concurrent existence of LABs and HABs and the elongated deformed grains confirm DRV mechanism during FSP of AISI D2. Beside this, it is well known that work hardening due to dislocation multiplication and recovery due to dislocation rearrangement are two competitive phenomena during hot working providing the steady state condition, dynamic equilibrium, reflecting equiaxed subgrain formation. Therefore, considering the highest amount of HABs and the equiaxed grain formation of Fig. 3(g) and Table 1, it can be perceived that the steady state condition was obtained in FSP at 400 rpm rotation rate. Whereas few amounts of HABs in the EBSD misorientation maps of Fig. 3(h), (i) and Table 1, it can be implied that during FSP conditions at rotation rates of 500–800 rpm, work hardening must be predominant phenomenon than work softening [22,34]. As shown in Fig. 6, the large non-deformable carbide particles (>1 m) are surrounded with the fine grains. These areas associating with a zone of lattice rotation, which is due to accumulation of the concentrated stress, are known as the particle-stimulated nucleation (PSN) and thereby CDRX mechanism. Sato et al. [35] have also reported the nucleation (CDRX mechanism) of the new grains during FSW of precipitation-hardened 6063 Al. It is well known that grain subdivision is essentially controlled by rotation rate of different parts of the same original grains towards different texture components [30]. Hence it is needed to study texture development in the FSPed samples. Fig. 7(a)–(d) shows {1 1 0} pole figures obtained from EBSD analysis of the FSPed specimens. In order to evaluate the preferred texture, the positions of ideal bcc shear texture components are elucidated in Fig. 7(e). In Fig. 7(a)–(d), the almost symmetric simple shear texture, which is characteristic of DRX attained from severe plastic deformation of FSW/P [36], is clearly apparent. This figure also reveals that the progressive variation of the texture intensity is accompanied with increasing the rotation rate of FSP in such a way that from rotation rate of 400–800 rpm goes along with monotonic increase
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
155
Fig. 7. (1 1 0) pole figures showing the bcc ferrite texture evolution as a function of FSP rotation rates of (a) 400, (b) 500, (c) 600, (d) 800 rpm and (e) the ideal bcc shear texture components. Note that each pole figure has its own scale bar and provides a qualitative observation of the texture evolution. In all pole figures the North Pole denotes the shear plane normal and east the shear plane direction as RD and TD of the SZs.
strain caused by the tool. It can be resulted that increasing the rotation rate of FSP, introducing higher strain and strain rate to the material, is associated with high number of independent slip systems activation, resulting in an increase of the texture intensity [36]. Furthermore, since the recovery processes involve a shortrange interaction between dislocations and subgrain boundaries, or between adjacent boundaries, they may lead to higher sliding on crystallographic planes and rotation of the lattice so that higher amount of the planes and directions are aligned towards the preferred directions (SDs) (preserving geometrical characteristics of plastic deformation) [36]. Therefore, a sharpening of deformation textures and a higher intensity of these deformation texture components appear, as given in Fig. 7(a)–(d), through increasing the rotation rate of FSP.
By increasing the FSP rotation rate, a clear systematic rotation of the texture can be observed about ND in such a way that in 400 rpm pole figures, the 1 1 0 poles are along with the RD (i.e. the distribution of the most of the (1 1 0) planes are parallel to RD). In the rotation rate of 500 rpm pole figures, the 1 1 0 poles are concentrated along with the TD (i.e. most of the (1 1 0) planes are perpendicular to RD). In the rotation rate of 600 rpm, the orientation 1 1 0 is the same as that of the 400 rpm but in upside down direction. The 1 1 0 poles of the rotation rate of 800 rpm are located by 45◦ in counter-close-wise direction with respect to RD as shown in Fig. 7(d). The systematic rotation of texture was previously shown across the advance through retreating sides of the FSW of 2519 Al alloy by Fonda and Bingert [37], and in FSW of AA6082-T651 by Ahmed et al. [38]. The dominant texture
156
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
Fig. 8. Orientation distribution function (ODF) of the SZs of the FSPed samples as a function of tool rotation rates. The ideal shear texture components of bcc metals and the custom prominent texture fibres.
components of all FSPed samples as indicated by the related pole figures in Fig. 7(a)–(d) mostly include D1, D2 which come from usual shear texture of rolling and recrystallization of bcc metals from Table 2 [39,40]. In this regard, Liu et al. [29] showed that the shear component is typically formed with inhomogeneous shear deformation due to the FSP. With the intention of more precisely assessing the texture and FSP rotation rates on dominance of texture evolution, the orientation distribution function (ODF) was calculated from the three experimentally measured, pole figures for each SZ. The ODFs of all
Table 2 Main ideal orientations in simple shear deformation of bcc materials [49]. Notation
{h k l} [u v w]
D1 D2 E1 E2 J1 J2 F
¯ [1 1 1] (1 1 2) (1¯ 1¯ 2) [1 1 1] ¯ [1 1 1] (0 1 1) (0 1¯ 1) [1 1 1] (0 1¯ 1) [2¯ 1 1] (1 1¯ 0) [1¯ 1¯ 1] (1 1 0) [0 0 1]
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
157
Table 3 EPMA chemical composition of AISI D2 elements in BM and FSPed sample. Element process BM FSP 600 rpm
Fig. 9. Orientation density along the different fibres versus ˚ and ϕ1 Euler angles, as a function of FSPed samples rotation rates.
FSPed samples are shown in Fig. 8 and the texture components and their intensities can be identified. In ODFs of all FSPed samples, no continuous bands of orientations extend in the 2 direction. Fig. 9(a)–(c) shows the prevalence fibres variations consisting of ␣, ␥ and which are appearing during FSP with different rotation rates. According to Fig. 9(a), the intensity of ␣-fibre (1 1 0//RD) is decreased by increasing the FSP rotation rates from 400 to 800 rpm. The homogeneous orientation density variation along the ␥-fibre ((1 1 1)//ND) in FSPed samples of 400–600 rpm and high increment of this fibre in 800 rpm represents the so-called pancake formation model of grains. In addition, the -fibre (0 1 1//TD) is decreased by rotation rates. The 400 and 500 rpm, because of the higher ␣-fibre compared to those of 600 and 800 rpm, show more random texture. The 800 rpm with the highest ␥-fibre (including {1 1 1} 1 2 3 component) growth by consuming the ␣-fibres with respect to those
V
Cr
0.74
7.95 8.45
Mn
Fe
Mo
0.27
81.20 86.66
0.61
of 400 and 500 rpm and therefore processes high stored energy components [41]. Considering the above results, it can be also implied that, at lower rotation rate of 400 rpm with the lower strain/strain rate, the lower slip system activated just provides the newly fine equiaxed grains by CDRX. At rotation rate of 500 rpm, according to crystal plasticity theory [42], the enough strain/strain rate or strain-induced progressive rotation of subgrain boundaries provides suitable slip system to be activated. According to Taylor’s original model defined elsewhere [37], it can be realized that if the dominant shear texture components satisfies the higher critical resolved shear stress, to minimize the total internal strain while maintaining compatible strain between grains in a polycrystalline structure, the lowest Taylor factor (M) will be obtained. Indeed, the different formed textures during various FSP rotation rates caused different Taylor factor. In this regard, in 500 rpm, the ␣-fibre decreased or the {1 1 1} 1 1 0 developed significantly and resulted in continuous lattice alignment in such a way that the high CPP/Ds to PSN and consistent with SD of the ideal simple shear reference frame. By increasing the rotation rate from 500 to 800 rpm, the Taylor factor increases. These results are not accompanied with the grain refinement trend obtained for different rotation rates of 400–800 rpm, while they are directly accompanied with the nanohardness measurements of 683, 1207, 552 and 583 HV for rotation rates of 400, 500, 600, and 800 rpm, accordingly. The predominant effect of texture with respect to microstructure refinement proved by Hall–Pech equation was previously reported by Wang et al. [43] in FSP of Mg–Al–Zn alloy. This also confirms the results obtained from IPFs of Fig. 3(d)–(f). The perpendicular oriented {1 1 0} planes to RD in 500 rpm FSPed samples indicate that in this condition, the CPPs are extensively affected by the applied axial pressure of the shoulder and backing plate [44] and consequently was higher than that of 400 and 600 rpm. On the other hand, it can be implied that at 500 rpm, the CPPs with the highest Young’s modulus is another responsible factor for better mechanical properties of this sample compared with the other FSPed samples [45]. As shown by EBSD results (Fig. 3(d)–(f)), the increment maintenance of the shear-type deformation texture with rotation rates confirms that DRV is the predominant mechanism of FSP. Alternatively, the EPMA results of Table 3, indicating the increment presence of the interstitials in ferrite compared with that of the BM and providing a significant drag on dislocations, results in positive strain rate sensitivity. In this regard, increasing the tool rotation rate provides more interstitial elements diffusion as well as a much stronger texture at rotation rate of 800 rpm compared to those obtained at the rotation rates of 400–600 rpm. Previous works [46] also showed that FSW/P microstructural trend seems to be like control rolled processing in which interstitial elements have a strong effect on texture formation during warm rolling. In conclusion, the positive strain rate sensitivity leads to more heterogeneous deformation and results in a sharpened texture which was shown in Fig. 7(a)–(d). Furthermore, when slip contributions to strain are limited by restricted numbers of slip systems or by solute drag effects, and boundary migration is similarly retarded by solute, lattice rotation may develop progressively at grain boundaries as dislocations accumulate during straining. This may lead to continuous dynamic recrystallization (CDRX) or extended recovery during FSP of AISI D2.
158
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159
particles is proposed as another mechanism during FSP. The EBSD’s analysis and the texture studies of samples indicate that the dominant source of the residual stress is plastic deformation. In 500 rpm with high amount of close-packed planes show the most plastic deformation and compressive residual stress. This sample with the {1 1 1} 1 1 0, {1 1 2} 1 1 1 and {1 1 0} 0 0 1 shear components and the lowest Taylor factor shows the highest nanohardness measurements. Therefore, the formed textures play more predominant effect on mechanical properties than grain refinement.
References
Fig. 10. Longitudinal and shear residual stresses along the processed direction at the surfaces of BM and FSPed samples at different rotation rates.
The longitudinal residual stresses of FSPed SZs shown in Fig. 10 lie in a range of 2–12% of 0.2% offset yield strength of the parent metal (illustrated in Guu’s work [47]). These values are somewhat smaller than those reported in previous literatures [20,48] of (20–70%), which were typically close to that of the dynamically recrystallized SZs. This inconsistency can be ascribed to the residual stress resource. As shown in Fig. 3(d)–(f), the elongated appearance of the new grains confirms that grain refining of the FSPed samples is more influenced by severe plastic deformation compared with the concurrent produced heat. In fact, the lower heat input or predominant stress state lead to such small residual stress maintenance in all FSPed samples. From 400 to 500 rpm rotation rate, the longitudinal residual stress has altered from tensile to compressive. However, from 500 to 600 and 800 rpm, it changed to higher level of tensile residual stresses. This is in good agreement with those obtained by Lombard et al. [48] in FSW of AA5083 and H321 alloy. The sharp reduction of tensile residual stress can be related to concurrent effect of the severe plastic deformation and generated heat of FSP at 500 rpm rotation rate respect to that of 400 rpm. Indeed, in FSPed region of FSPed sample at rotation rate of 500 rpm as previously indicated by EBSDs’ and texture studies, the highly dense CPPs in accompany with the usual shear slip system of bcc metals reduced the residual stress of this FSPed sample. This residual stress decrease can also be ascribed to this situation which satisfies the Taylor’s original model. In contrast, it is possible to conclude that in FSPed samples at rotation rates of 600 and 800 rpm, the higher plastic deformation existence (with higher LABs/dislocation density) led to increase of the longitudinal residual stresses to higher level. Continuous increase of the shear stress of Fig. 10 infers the ascending trend of severe plastic deformation with respect to FSP rotation rates. 4. Conclusions In the present study, crystallographic texture development and the mechanisms happening during FSP of AISI D2 steel have been studied. The EBSD analysis shows that DRV and CDRX are the predominant mechanisms of grain refinement. Decreasing the HABs with increasing the rotation rates implies that work hardening or DRV must be the predominant phenomenon than work softening. This can be due to (I) an increased amount of solute carbon in ferrite retarding the formation and migration of HABs, and (II) the increased trend of residual stress by rotation rates during FSP which indicate the increase trend of dislocation densities. In addition to DRV, CDRX caused by PSN of large non-shearable carbide
[1] W.A. Baeslack, K.V. Jata, T.J. Lienert, Structure, properties and fracture of friction stir welds in a high-temperature Al–8.5Fe–1.3V–1.7Si alloy (AA-8009), J. Mater. Sci. 41 (2006) 2939–2951. [2] V. Balasubramanian, Relationship between base metal properties and friction stir welding process parameters, Mater. Sci. Eng. A 480 (2008) 397–403. [3] M. Dumont, A. Steuwer, A. Deschamps, M. Peel, P. Withers, Microstructure mapping in friction stir welds of 7449 aluminium alloy using SAXS, Acta Mater. 54 (2006) 4793–4801. [4] A.K. Lakshminarayanan, V. Balasubramanian, An assessment of microstructure, hardness, tensile and impact strength of friction stir welded ferritic stainless steel joints, Mater. Des. 31 (2010) 4592–4600. [5] S. Noh, R. Kasada, A. Kimura, S.H.C. Park, S. Hirano, Microstructure and mechanical properties of friction stir processed ODS ferritic steels, J. Nucl. Mater. 417 (1) (2011) 245–248. [6] P.J. Hurley, F.J. Humphreys, The application of EBSD to the study of substructural development in a cold rolled single-phase aluminium alloy, Acta Mater. 51 (2003) 1087–1102. [7] R.W. Fonda, J.F. Bingert, Precipitation and grain refinement in a 2195 Al friction stir weld, Metall. Mater. Trans. 37 (2006) 3593–3594. [8] H. Fujii, L. Cui, M. Maeda, K. Nogi, Effect of tool shape on mechanical properties and microstructure of friction stir welded aluminum alloys, Mater. Sci. Eng. A 419 (2006) 25–31. [9] H. Fujii, L. Cui, N. Tsuji, M. Maeda, K. Nakata, K. Nogi, Friction stir welding of carbon steels, Mater. Sci. Eng. A 429 (2006) 50–57. [10] R. Ueji, H. Fujii, L. Cui, A. Nishioka, K. Kunishige, K. Nogi, Friction stir welding of ultrafine grained plain low-carbon steel formed by the martensite process, Mater. Sci. Eng. A 423 (2006) 324–330. [11] S.H.S. Park, Y.S. Sato, H. Kokawa, K. Okamoto, S. Hirano, M. Inagaki, Rapid formation of the sigma phase in 304 stainless steel during friction stir welding, Scr. Mater. 49 (2003) 1175–1180. [12] A. Reynolds, Structure, properties, and residual stress of 304L stainless steel friction stir welds, Scr. Mater. 48 (2003) 1289–1294. [13] X.K. Zhu, Y.J. Chao, Numerical simulation of transient temperature and residual stresses in friction stir welding of 304L stainless steel, J. Mater. Process. Technol. 146 (2004) 263–272. [14] W. Woo, H. Choo, D. Brown, S. Vogel, P. Liaw, Z. Feng, Texture analysis of a friction stir processed 6061-T6 aluminum alloy using neutron diffraction, Acta Mater. 54 (2006) 3871–3882. [15] G. Buffa, L. Fratini, S. Pasta, R. Shivpuri, On the thermo-mechanical loads and the resultant residual stresses in friction stir processing operations, CIRP Ann. – Manuf. Technol. 57 (2008) 287–290. [16] O. Hatamleh, I.V. Rivero, A. Maredia, Residual stresses in friction-stir-welded 2195 and 7075 aluminum alloys, Metall. Mater. Trans. 39 (2008) 2867–2874. [17] C.M. Hu, C.M. Lai, P.W. Kao, N.J. Ho, J.C. Huang, Solute-enhanced tensile ductility of ultrafine-grained Al–Zn alloy fabricated by friction stir processing, Scr. Mater. 60 (2009) 639–642. [18] L. Johannes, R. Mishra, Multiple passes of friction stir processing for the creation of superplastic 7075 aluminum, Mater. Sci. Eng. A 464 (2007) 255–260. [19] N. Yasavol, A. Abdollah-zadeh, D.M. Rodrigues, Microstructure and mechanical properties of friction stir processed AISI D2 tool steel, in: Second International Conference on Friction Stir Welding and Processing, 26–27th January, 2012. [20] W. Woo, Z. Feng, X.-L. Wang, S.A. David, Neutron diffraction measurements of residual stresses in friction stir welding: a review, Sci. Technol. Weld. Join. (2004) 23–32. [21] Y. Sato, H. Kokawa, K. Ikeda, M. Enomoto, S. Jogan, T. Hashimato, Microtexture in the friction-stir weld of an aluminum alloy, Metall. Mater. Trans. 32 (2008) 941–947. [22] K. Dehghani, A. Chabok, Dependence of Zener parameter on the nanograins formed during friction stir processing of interstitial free steels, Mater. Sci. Eng. A 528 (2011) 4325–4330. [23] W. Xu, J. Liu, G. Luan, C. Dong, Temperature evolution, microstructure and mechanical properties of friction stir welded thick 2219-O aluminum alloy joints, Mater. Des. 30 (2008) 1886–1893. [24] R.S. Mishra, M.W. Mahoney, Friction Stir Welding and Processing, ASM International, United States of America, 2007. [25] Ø. Frigaard, Ø. Grong, T. Mildingo, A process model for friction stir welding of age hardening aluminum alloys, Metall. Mater. Trans. 32A (2001) 1189–1200. [26] A. Gerlich, M. Yamamoto, T.H. North, Strain rates and grain growth in Al 5754 and Al 6061 friction stir spot welds, Metall. Mater. Trans. 38 (2007) 1291–1302.
N. Yasavol et al. / Applied Surface Science 293 (2014) 151–159 [27] A.P. Gerlich, T. Shibayanagi, Grain boundary sliding during friction stir spot welding of an aluminum alloy, Scr. Mater. 60 (2009) 236–239. [28] F.J. Humphreys, Characterisation of fine-scale microstructures by electron backscatter diffraction (EBSD), Scr. Mater. 51 (2004) 771–776. [29] F.C. Liu, Z.Y. Ma, L.Q. Chen, Low-temperature superplasticity of Al–Mg–Sc alloy produced by friction stir processing, Scr. Mater. 60 (2009) 968–971. [30] S. Mironov, Y.S. Sato, H. Kokawa, Microstructural evolution during friction stirprocessing of pure iron, Acta Mater. 56 (2008) 2602–2614. [31] R.S. Coelho, A. Kostka, S. Sheikhi, J. dos Santos, A.R. Pyzalla, Microstructure and mechanical properties of an AA6181-T4 auminium alloy to HC340LA high strength steel friction stir overlap weld, Adv. Eng. Mater. 10 (2008) 961–972. [32] K.A.A. Hassan, A.F. Norman, D.A. Price, P.B. Prangnell, Stability of nugget zone grain structures in high strength Al alloy friction stir welds during solution treatment, Acta Mater. 51 (2003) 1923–1936. [33] T.J. Lienert, W.L. Stellwag JR., B.B. Grimmett, R.W. Warke, Friction stir welding studies on mild steels, Weld. J. (2003) 1–9. [34] Y. Sato, H. Yamanoi, H. Kokawa, T. Furuhara, Microstructural evolution of ultrahigh carbon steel during friction stir welding, Scr. Mater. 57 (2007) 557–560. [35] Y. Sato, M. Urata, H. Kokawa, Parameters controlling microstructure and hardness during friction-stir welding of precipitation-hardenable aluminum alloy 6063, Mater. Sci. Eng. A 33 (2002) 625–635. [36] U.F. Kocks, C.N. Tomé, H.-R. Wenk, Texture and Anisotropy Preferred Orientations in Polycrystals and their Effect on Materials Properties, Cambridge University Press, United Kingdom, 1998. [37] R.W. Fonda, J.F. Bingert, Microstructural evolution in the heat-affected zone of a friction stir weld, Metall. Mater. Trans. 35 (2004) 1487–1499. [38] M.M.Z. Ahmed, B.P. Wynne, W.M. Rainforth, P.L. Threadgill, Quantifying crystallographic texture in the probe-dominated region of thick-section friction-stir-welded aluminium, Scr. Mater. 59 (2008) 507–510.
159
[39] I.L. Dillamore, W.T. Roberts, Rolling textures in FCC and BCC metals, Acta Metall. 12 (1964) 282–294. [40] M. Hoscher, D. Raabe, K. Lucke, Rolling and recrystallization textures of bcc steels, Steel Res. 62 (1991) 567–575. [41] B. Verlinden, J. Driver, I. Samajdar, R. Doherty, in: R.W. Cahn (Ed.), Thermo-Mechanical Processing of Metallic Materials, Elsevier, Amsterdam, 2007. [42] B. Heinz, B. Skrotzki, Characterization of a friction-stir-welded aluminum alloy 6013, Metall. Mater. Trans. 33 (2002) 497–504. [43] Y.N. Wang, C.I. Chang, C.J. Lee, H.K. Lin, J.C. Huang, Texture and weak grain size dependence in friction stir processed Mg–Al–Zn alloy, Scr. Mater. 55 (2006) 637–640. [44] S. Park, Y. Sato, H. Kokawa, Basal plane texture and flow pattern in friction stir weld of a magnesium alloy, Mater. Sci. Eng. A 34 (2003) 987–997. [45] W.D.J. Callister, Materials Science and Engineering: An Introduction, John Wiley and Sons, New York, 1940. [46] R. Nandan, T. DebRoy, H.K.D.H. Bhadeshia, Recent advances in friction-stir welding – process, weldment structure and properties, Prog. Mater. Sci. 53 (2008) 980–1023. [47] Y. Guu, AFM surface imaging of AISI D2 tool steel machined by the EDM process, Appl. Surf. Sci. 242 (2005) 245–250. [48] H. Lombard, D.G. Hattingh, A. Steuwer, M.N. James, Effect of process parameters on the residual stresses in AA5083-H321 friction stir welds, Mater. Sci. Eng. A 501 (2009) 119–124. [49] S. Lia, J.I. Beyerleinb, M.A.M. Bourkea, Texture formation during equal channel angular extrusion of fcc and bcc materials: comparison with simple shear, Mater. Sci. Eng. A 39 (2005) 66–77.