The effect of temperature on strain-rate sensitivity in

2 downloads 0 Views 154KB Size Report
the tested AlMg6 alloy sheet the SRS is negative and decreasing type around the room ..... Aluminium and Magnesium for Automotive Applications, Cleveland,.

Journal of Materials Processing Technology 125±126 (2002) 193±198

The effect of temperature on strain-rate sensitivity in high strength Al±Mg alloy sheet E. Romhanjia,*, M. Dudukovskab, D. GlisÏicÂa a

Faculty of Technology and Metallurgy, University of Belgrade, Kaarnegijeva 4, P.O. Box 3503, 1120 Belgrade, Yugoslavia b Faculty of Metallurgy, University of Skopje, FRY, Macedonia Received 15 December 2001; accepted 23 February 2002

Abstract Comprehensive experimental work performed to assess the temperature and strain-rate affected deformation behaviour of commercial AlMg6 type, 1.0 mm thick annealed sheet. Elevated temperature tension test performed at temperatures ranged from 25 to 300 8C, and three cross-head rates, giving initial strain rates of e_ 1 ˆ 6:7  10 4 s 1 , e_ 2 ˆ 6:7  10 3 s 1 and e_ 3 ˆ 6:7  10 2 s 1 . Steady state strain-rate sensitivity (SRS) parameters were calculated (m ˆ d ln s=d ln e_ ) for the strain-rate ratios of e_ 1 =_e2 (1:10), e_ 2 =_e3 (10:100). It was shown that in the tested AlMg6 alloy sheet the SRS is negative and decreasing type around the room temperature due acting the dynamic strain ageing (DSA). It becomes positive at higher temperatures when the DSA weakening. This transition temperature increases by strain rate. The monotonic increase of the SRS brought by temperature is assumed to be due the enhancement of dynamic recovery and it appeared to be independent on strain rate, except at 300 8C. At 300 8C decreasing the strain rate brings considerable increase of the SRS. The experienced increase of the terminal m values from 0.15 to 0.4 for the applied strain-rate ratios of e_ 1 =_e2 (1:10) and e_ 2 =_e3 (10:100), respectively, is assumed to be the result of changing the deformation mechanism from recovery to diffusion-controlled solute drag. However, the performed research has shown that the attained increase of the SRS was not followed with appropriate ductility improvement. Analysis of the strain localization (necking) has shown that increasing the strain rate from 6:7  10 4 to 6:7  10 2 s 1, the temperature brought softening can be compensated, and the process of strain localization (necking) shifted to higher strains. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Al±Mg alloy; Strain-rate sensitivity; Temperature; Strain rate; Necking

1. Introduction The elevated temperature deformation behaviour of metals and alloys is of continuing scienti®c and commercial interest due the bene®ts of extended formability. The Al±Mg type alloy sheets are of interest because, besides the good weldability or corrosion resistance, their room temperature stretching performance is rather limited [1]. Important ductility improvement can be achieved in warm forming [2,3], especially in coarse grained and low impurity Al±Mg alloys [4±6], when the ductility even approaches the superplastic behaviour without special structure design and deformation conditions. One of the general observation related to warm temperature deformation is that the amount of uniform deformation decreases with increase of temperature, while the ductility improvement is due to the extended range of post-uniform elongation, controlled by the strain-rate hardening properties [2,3,6]. In other words, by increasing the temperature, the work hardening decreases, while the strain *

Corresponding author.

rate in¯uenced hardening becomes more effective. The improved stretchability at elevated temperatures derives from the improved necking resistance, brought about by strain-rate sensitivity (SRS) in¯uenced strengthening. From the scienti®c viewpoint still there is a lack of understanding of the interactions of different factors in controlling the SRS, strain hardening, grain boundary sliding, cavitations, etc. [2±8]. Also, it is important to identify the lowest suitable warm forming temperature in order to avoid the excessive oxidation or internal cavity development during stretching [9]. The aim of the present work is to show some results concerning the temperature and strain-rate affected variations of deformation behaviour of the AlMg6 type alloy sheet. 2. Material The as received AlMg6.8 sheet was 3.0 mm thick, in the annealed condition with the chemical composition in wt.%, of Al±6.8 Mg±0.51 Mn±0.1 Si±0.2 Fe±0.03 Zn±0.05 Ti. It was further cold rolled to 1.0 mm and annealed at 320 8C for

0924-0136/02/$ ± see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 4 - 0 1 3 6 ( 0 2 ) 0 0 3 0 8 - 4

194

E. Romhanji et al. / Journal of Materials Processing Technology 125±126 (2002) 193±198

3 h, in an inert gas atmosphere. Metallographic examinations revealed a ®ne grained homogenous microstructure with the average grain diameters ranged to 30 mm. 3. Tensile testing Tension tests, in the temperature range from 25 to 300 8C, were carried out on ``Instron'' tensile testing machine, using specimens with 110 mm gauge length and 20 mm widths. The testing temperatures were controlled using the thermocouple attached to the central part of gauge lengths. Three cross-head rates were applied, giving initial strain rates of e_ 1 ˆ 6:7  10 4 s 1 , e_ 2 ˆ 6:7  10 3 s 1 and e_ 3 ˆ 6:7  10 2 s 1 . SRS parameters, m ˆ d ln s=d ln e_ , were calculated over the entire tensile strain range. Due the necking effect at higher temperatures (200, 250 and 300 8C), additional samples were strained up to different levels and the area of minimum cross sections were measured, allowing the correction of true stress values. So, besides the length strain eL (assuming homogenous straining: the sample sides supposed to remain parallel over the whole range of deformation), the area strain (contraction), eC, in the minimum cross section was also calculated. As a convenience in expressing the deformation homogeneity or the localization intensity the eC/eL ratio was assumed. Comparing the recorded elongations and the measured total elongations on fractured samples, the proportionality factor was established between the strain in the gauge length and the cross-head travel. 4. Results 4.1. Strain-rate sensitivity The SRS parameter m variation during uniaxial stretching at different temperatures is shown in Fig. 1(a) and (b), for two

Fig. 1. The SRS (m) variation during straining (e) at different temperatures (25±300 8C) and strain-rate pairs: (a) e_ 2 =_e1 ˆ 10:1; (b) e_ 3 =_e2 ˆ 100:10.

different strain-rate ratios. The temperature effect is recognized through the clear increase of the m values for both strain-rate ratios. Around the room temperature (25 and 50 8C or even at 100 8C for the case in Fig. 1(b)) the m values are negative (become negative after e ˆ 0:05) and moderately decrease as the strain increases, while at higher temperatures they are positive and increasing. The total range of m values for e_ 2 =_e1 ˆ 10:1 are larger (Fig. 1(a)), than for e_ 3 =_e2 ˆ 100:10 (Fig. 1(b)), i.e. for the same strain ratio in range of one order higher rates. This difference is mainly brought by the change in m value at 300 8C, i.e. at all the lower temperatures the strain-rate effect is negligible. It is clearly shown in Fig. 2 through the variation of terminal values of the SRS parameters mterm taken from Fig. 1. The mterm values

Fig. 2. Temperature influence on the terminal values of SRS (mterm) for the two applied strain-rate ratios.

E. Romhanji et al. / Journal of Materials Processing Technology 125±126 (2002) 193±198

195

Fig. 3. Temperature influence on strength after different strain rates applied in the tension test: (a) yield strength (UTS); (b) ductility (et and eh).

were found to increase monotonically with temperature. The strain-rate ratio effect was recognized only at 300 8C. 4.2. Mechanical properties The basic temperature effect on strength of the tested AlMg6 type alloy sheet appeared as an intensive lowering of the tensile strength (UTS) over the entire strain range (Fig. 3(a)), and a somewhat slower drop of the yield stress (YS), rather evident at higher temperatures (t > 100 8C). So, in the temperature range of 25±300 8C, and after employing strain rates of e_ 1 ˆ 6:7  10 4 s 1 , e_ 2 ˆ 6:7  10 3 s 1 and e_ 3 ˆ 6:7  10 2 s 1 , the UTS dropped 71, 61 and 58%, respectively. The appropriate YS change was in a same manner but less pronounced: 58, 32 and 25%. Up to 100 8C, the strain rate in¯uence on the mechanical properties is weak, although the UTS are clearly lower at higher strain rates (Fig. 3(a)). After the inversion, which occurred at about 75 8C, the UTS strongly differentiated by the applied strain rates, giving higher UTS values for higher strain rates

at the entire temperature range of t > 100 8C. The total elongation (et) is lower at higher strain rates, while a rather slowly increasing trend is brought by temperature (Fig. 3(b)). On the other hand, the homogenous elongation (eh) strongly decreases with a rise in the temperature, but the strain-rate effect can be recognized at t  200 8C. 4.3. Strain localization Observation of the strain localization at 200, 250 and 300 8C, for the three strain rates, e_ 1 and e_ 2 (Fig. 4(a) and (b)), shows that the strain localization (eC/eL) becomes more intensive with increasing the temperature. This effect vanished after the highest strain rate, e_ 3 , was applied (Fig. 4(c)), and for all three temperatures a rather common and distinct two stages of strain localization were recognized: (i) The early stage, up to eC  0:12, which appeared to be linear and approximately the same for all strain rates, with eC =eL ˆ 1.

196

E. Romhanji et al. / Journal of Materials Processing Technology 125±126 (2002) 193±198

These stages of localization were recognizable even at lower strain rates e_ 2 and e_ 1 , when the tests conducted at 200 8C (Fig. 4(a) and (b)). At higher temperatures, for these two strain rates, the ®rst stage seems to be suppressed, as it is visible for the combinations e_ 1 and 250 8C (Fig. 4(a)) or e_ 2 and 300 8C (Fig. 4(b)). So, the localization proceeds almost from the beginning with continuously increasing intensity, in a rather parabolic manner. 5. Discussion

Fig. 4. Length strain (eL)±contraction (eC) dependence on temperature for strain rates of: (a) e_ 1 ˆ 6:7  10 4 s 1 ; (b) e_ 2 ˆ 6:7  10 3 s 1 ; (c) e_ 3 ˆ 6:7  10 2 s 1 .

(ii) Intensive localization, when the necking process becomes more pronounced with eC =eL  2:5, i.e. the deformation proceeds mainly within the neck.

The low temperature SRS (at temperatures 25±100 8C) is negative, and it is a common feature of the alloys characterized by the appearance of dynamic strain ageing (DSA) [10± 12]. It is believed [13±15] that the m value is a result of dislocation±solute interaction, stated as a solute atom composition variation at arrested dislocations, due the change of strain rate [16]. Brie¯y, at higher strain rates, the solute drag is weaker and the increased amount of mobile dislocations should be accompanied with a lower ¯ow stress. Conversely, at lower strain rates the ¯ow stress increases. The m values during straining can be considered as a consequence of the interplay between: (i) strain hardening due to the increase of dislocation density with deformation, which tends to move m toward positive values, and (ii) DSA due to the increase the number of obstacles and deformation induced vacancies, which facilitate diffusion of Mg solute, lowering m towards negative values. A moderate loss of strain-rate hardening ability (m becomes more negative), brought about by straining, is also reported earlier [17], and it can be rationalized by the weakening of strain hardening contribution. All the considered SRS variation and explanations seem to be in a general agreement with the van den Beukel model [11,18] for predicting SRS. The SRS becomes less negative by increasing the temperature, and after reaching a certain temperature of about 50 8C in the range of lower strain rates (Fig. 1(a)) or 100 8C for higher strain rates (Fig. 1(b)), the m values become positive for the entire strain range. This transition seems to coincide with the temperature range when the DSA vanishes in the tested material (at least the serrations as its macroscopic manifestation) [19]. Increasing the temperature further (up to 300 8C), as well as by straining, the m values increase in a same manner for both strain rates, what is assumed to be due the gradual progress of the dynamic recovery process. Previously, it was stated [20] that at elevated temperatures the ¯ow stress always re¯ects a balance between hardening and softening processes, where the softening process is a variable part of the total hardening. Accordingly, it is assumed that the monotonic decrease of the uniform elongation with temperature (Fig. 3(b)) is a manifestation of the gradual development of dynamic recovery what was shown to be the dominant deformation mechanism at moderate temperatures [19,21]. It should be emphasized that the noticed m variations are not fully

E. Romhanji et al. / Journal of Materials Processing Technology 125±126 (2002) 193±198

understood regarding the details of the dislocation/dislocation or dislocation/solute reactions, although the increase of the m values has been correlated with the increased dislocation density in cell walls brought by dynamic recovery [2]. Really, the higher density of mobile dislocations is necessarily followed by additional hardening, recognized as the increase of UTS values (Fig. 3(a)). The less pronounced strain rate induced change of the YS is a consequence of the small difference in the m values at early stage of straining. Further, it was considered [22,23] that in warm deformation, high m values in Al±Mg alloys appear when viscous glide (in which dislocations move together with the solute atmospheres) come into play, in parallel with the dynamic recovery. The distinctly higher m values experienced over the entire strain range at 300 8C, in the low strain-rate test with e_ 1 =_e2 ˆ 1:10, and when the mterm is 0.4 (Figs. 1(a) and 2), is a case when probably the diffusion-controlled solute drag governs. This assumption is based on the earlier established deformation mechanism change from recovery to diffusion-controlled solute drag in Al±Mg-5182 type alloy at the same temperature±strain-rate conditions [24]. Flattening of the true stress±strain curves also follows those changes, what is recognized in the performed tension test at 300 8C with the lowest strain rate of e_ 1 ˆ 6:7  10 4 s 1 [19]. For the strain-rate ratio of e_ 2 =_e3 ˆ 10:100 such a behaviour is less obvious, which is explicitly shown by the drop of mterm value in Fig. 2. In spite of the high m values attained for low strain rates and higher temperatures (specially at 300 8C), the ductility improvement was limited (Fig. 3(b)). In other words, the attained ductility is far from any quasi-superplastic behaviour as in some high purity Al±Mg alloys [5]. Having in mind the possible strong in¯uence of the cavitations in the tested commercial grade alloy, it could be supposed that in the same alloy with lower impurity content, the ductility could be probably improved. The detrimental strain-rate effect on elevated temperature ductility in the tested alloy is not clear, because, except in the mentioned case of the decrease in m with strain rate at 300 8C, the strain-rate effect on the terminal m values is negligible (Fig. 2). The understanding of the strain-rate effect should be related to the ®ndings that the dynamic recovery is suppressed by increasing the strain rate [21]. The assumption that the cavitations tendency is lowered by increasing the strain rate [25,26] should be further cleared up by taking into account structural features in¯uences. The strain localization (necking) at 250 and 300 8C start from the very beginning of straining, after applying lower strain rates of e_ 1 ˆ 6:7  10 4 s 1 , and e_ 2 ˆ 6:7  10 3 s 1 (Fig. 4(a) and (b)). The strain hardening and the strain-rate hardening are obviously not enough to keep the deformation homogenous. Increasing the strain rate to e_ 3 ˆ 6:7  10 2 s 1 the temperature induced softening seems to be compensated, and the process of strain localization is shifted to higher strains independently of the temperature. The temperature effect appeared to be completely suppressed

197

at e_ 3 ˆ 6:7  10 2 s 1 (Fig. 4(c)), while at 200 8C, the region of homogenous deformation …eC =eL ˆ 1† is independent of the strain rate (Fig. 4(a)±(c)). 6. Summary Elevated temperature tension tests have been performed in the temperature range from 25 to 300 8C, and three initial strain rates of e_ 1 ˆ 6:7  10 4 s 1 , e_ 2 ˆ 6:7  10 3 s 1 and e_ 3 ˆ 6:7  10 2 s 1 . Steady state SRS parameters were calculated m ˆ d ln s=d ln e_ comparing the stress±strain curves obtained for the strain-rate ratios of e_ 1 =_e2 (1:10), e_ 2 =_e3 (10:100). The SRS is a property that develops during straining. In the Al±Mg alloys the SRS is negative and of a decreasing nature around room temperature (25±100 8C) due to the action of DSA. It becomes positive and increasing at higher temperatures when the DSA is weaker. The transition from negative to positive values seems to be shifted to higher temperatures at higher strain rates. The monotonic increase of the SRS brought about by temperature is due the enhancement of dynamic recovery and not in¯uenced by strain rate, except at 300 8C. At 300 8C, a considerable increase of the SRS is brought about by decreasing the strain rate. The experienced increase of the terminal m values from 0.15 to 0.4 after applying strain-rate ratios of e_ 2 =_e3 (10:100) and e_ 1 =_e2 (1:10), respectively, is assumed to be the result of changing the deformation mechanism from recovery to diffusion-controlled solute drag. In spite of the high terminal m values attained at 300 8C for low strain rates (0.4), the ductility improvement was rather limited, emphasizing the importance of other structural characteristics which limit the elevated temperature ductility in the tested commercial grade AlMg6 type alloy sheet. Concerning this result, the cavitation tendency is probably the main source of the ductility limitation. On the other hand, increasing the strain rate from 6:7  10 4 to 6:7  10 2 s 1, the temperature brought softening can be compensated, and the process of strain localization (necking) shifted to higher strains. References [1] E. Romhanji, M. Popovic, D. Glisic, V. Milenkovic, J. Mater. Sci. 33 (1998) 1037. [2] R.A. Ayres, Metall. Trans. A 10 (1979) 849. [3] W.T. Roberts, D.V. Wilson, in: S. Blecic, H. Novi-Yugoslavia (Eds.), Proceedings of the International Symposium on Plasticity and Resistance to Metal Deformation (PLOD-IV), 1984, p. 163. [4] L. Ivanchev, D. Chavderova, in: S. Blecic, H. Novi-Yugoslavia (Eds.), Proceedings of the International Symposium on Plasticity and Resistance to Metal Deformation (PLOD-VI), 1986, p. 348. [5] E.M. Taleff, G.A. Henshall, D.R. Lesuer, T.G. Nieh, J. Wadsworth, in: J.D. Bryant, D.R. White (Eds.), Proceedings of the Conference on Aluminium and Magnesium for Automotive Applications, Cleveland, OH, TMS, Warrendale, PA, 1995, p. 125.

198

E. Romhanji et al. / Journal of Materials Processing Technology 125±126 (2002) 193±198

[6] E.M. Taleff, D.R. Lesuer, J. Wadsworth, Metall. Trans. A 27 (1996) 343. [7] I.C. Hsiao, S.W. Su, J.C. Huang, Metall. Trans. A 31 (2000) 2169. [8] D.J. Lloyd, Metall. Trans. A 11 (1980) 1287. [9] G. L'Esperance, W.T. Roberts, M.H. Loretto, D.V. Wilson, Proceedings of the 12th Biennal Congress IDDRG, S. Margerita Ligure, 1982, p. 97. [10] J.M. Robinson, M.P. Shaw, Mater. Sci. Eng. A 159 (1992) 159. [11] A. van den Beukel, Phys. Stat. Sol. (a) 30 (1975) 197. [12] E. Nes, Z. Metall. 69 (1978) 35. [13] E. Pink, A. Grinberg, Mater. Sci. Eng. 51 (1981) 1. [14] L.P. Kubin, A. Styczynski, Y. Estrin, Scripta Metall. Mater. 26 (1992) 1423. [15] E. Pink, in: P.O. Kettunen, et al. (Eds.), Proceedings of the Conference on Strength of Metals and Alloys (ICSMA 98), Vol. 1, Pergamon Press, Oxford, 1988, p. 494. [16] P.G. McCormick, in: A.K. Sachdev, J.D. Embury (Eds.), Proceedings of the Conference on Formability and Metallurgical Structure, TMS± AIME, Orlando, FL, 1986, p. 227.

[17] L.P. Kubin, Y. Estrin, Scripta Metall. 23 (1989) 815. [18] S.H. Van den Brink, A. van den Beukel, P.G. McCormick, Phys. Stat. Sol. (a) 30 (1975) 469. [19] M. Dudukovska, Ph.D. Thesis, University of Skopje, 1999. [20] U.F. Kocks, in: G. Kraus (Ed.), Proceedings of the Conference on Deformation, Processing and Structure, ASM, Metals Park, OH, 1982, p. 89. [21] R.A. Ayres, Metall. Trans. A 8 (1977) 487. [22] P. Yavari, T.G. Langdon, Acta Metall. 30 (1982) 2181. [23] F.A. Mohamed, T.G. Langdon, Acta Metall. 23 (1975) 117. [24] D.A. Hughes, M.E. Kassner, Mg. Stout, J.S. Vetrano, JOM 50 (6) (1998) 16. [25] R.E. Smallman, in: S. Blecic, H. Novi-Yugoslavia (Eds.), Proceedings of the International Symposium on Plasticity and Resistance to Metal Deformation (PLOD-IV), 1984, p. 9. [26] A.K. Ghosh, C.H. Hamilton, Proceedings of the Fifth International Symposium on Strength of Metals and Alloys, Vol. 2, Pergamon Press, Oxford, 1975, p. 905.