Effects of Microstructure on the Mechanical Properties and Stress Corrosion Cracking of an Al-Zn-Mg-Sc-Zr Alloy by Various Temper Treatments

Transcription

1 Materials Transactions, Vol. 48, No. 3 (2007) pp. 600 to 609 #2007 The Japan Institute of Metals EXPRESS REGULAR ARTICLE Effects of Microstructure on the Mechanical Properties and Stress Corrosion Cracking of an Al-Zn-Mg-Sc-Zr Alloy by Various Temper Treatments Ling-Mei Wu 1; * 1, Wen-Hsiung Wang 1; * 2, Yung-Fu Hsu 2 and Shan Trong 3 1 Department of Mechanical Engineering, National Taiwan University, Taipei, Taiwan 106, R. O. China 2 Department of Materials and Mineral Resources Engineering, National Taipei University of Technology, Taipei, Taiwan 106, R. O. China 3 Materials and Electro-Optics Research Division, Chung-Shan Institute of Science and Technology, Lung-Tan, Taiwan 325, R. O. China The high strength Al-Zn-Mg alloy used in the aerospace industry is strengthened by coherent G.P. zones and semicoherent 0 phase. However, this series of aluminum alloys are susceptible to stress corrosion cracking (SCC), particularly when aged to the peak-aged state of T6 temper. In this study, the effect of microstructure on mechanical properties and stress corrosion cracking of the alloy was investigated for alloys tempered to T6, RRA, Two-Step and T7 conditions by tensile test in air, SCC test and polarization test in a 3.5%NaCl solution. It is shown that the improvement in SCC resistance correlates very well with the size of matrix precipitates and grain boundary precipitates. T7 temper can produce larger sizes of both the matrix precipitates and grain boundary precipitates than that of T6, RRA and Two-step tempers, causing a decrease in the length and density of dislocation lines, which results in the decrease of stress concentration at grain boundary and greatly improving the SCC resistance. The addition of Sc and Zr to high strength Al-Zn-Mg alloys has been found to simultaneously improve the tensile strength and SCC resistance. [doi: /matertrans ] (Received October 2, 2006; Accepted January 16, 2007; Published February 25, 2007) Keywords: aging, stress corrosion cracking, grain boundary precipitates 1. Introduction High strength Al-Zn-Mg (7xxx series) aluminum alloys are widely used in airframe construction due to their high strength to density ratio. However, this series of alloys are susceptible to stress corrosion cracking (SCC) in the highest strength temper (T6). The susceptibility of alloys to SCC has been shown to strongly depend on their chemical composition and heat treatment. 1) Speidel 2) also indicated that the SCC susceptibility of the precipitation hardening aluminum alloy increased with aging time, and the extreme susceptibility was occurred under peak-aging condition. While at over-aging condition, the SCC susceptibility is reduced but with loss of mechanical strength about 10 15% comparing to the T6 condition. Consequently, the increase in SCC resistance of alloys was associated with a lower strength. It has been observed 3 6) that the main differences between the T6 and T7 tempers are related to the width of the precipitated free-zones, distribution of matrix precipitates, dislocation density, composition, size and distribution of grain boundary precipitates. These aspects of aluminum alloys are assumed to determine the susceptibility to SCC. Moreover, the sensitivity of stress corrosion for wrought commercial aluminum alloys at different aged stages has been studied and discussed. 7,8) A heat treatment known as retrogression and reaging (RRA) was proposed to reduce the SCC susceptibility in Al- Zn-Mg high strength aluminum alloys, while maintaining the strength of the T6 temper. 9) The RRA treatment has been applied to the alloy in the T6 temper and consists of reheating the alloy for a short time in the temperature range of 160 * 1 Graduate Student, National Taiwan University * 2 Corresponding author, whwang@ntu.edu.tw 220 C, followed by reaging using the original T6 temper. The main microstructural change after RRA temper is a very fine distribution of 0 precipitates in the matrix similar to T6 temper, and a larger size of precipitates on grain boundaries distributed similarly to T7 temper. 5,10,11) The combination microstructures could effectively improve both SCC resistance and strength. In addition, two-step aging treatments are often employed in high-strength 7xxx series aluminum alloys to obtained maximum precipitation hardening and better stress-corrosion resistance. 12) It is achieved through a twostep aging treatment, first aging at C for 4 8 h, followed by second aging at 150 C for 5 8 h. Two-step aging treatments have superior SCC resistance due to maintain the T6 temper microstructure and with a coarser precipitates along grain boundary. These specific aging treatments have a significant influence on the size, morphology, density and chemistry of precipitates and hence on the alloy properties in service. Furthermore, the density of fine precipitates which is responsible for the high strength and the SCC resistance of the alloy is generally thought to be controlled by the heat treatment. SCC damage occurs under dynamic loading in a corrosive environment. Several investigations have reported that the SCC mechanism involves anodic dissolution, 13 15) hydrogeninduced cracking (HIC), 5,13,16 18) passive film rupture, 13,19) hydrogen embrittlement (HE), 14,15,20) magnesium segregation to grain boundaries 21 23) and a precipitate free zone (PFZ) along grain boundary. 24,25) However, the microstructural characteristics of Al-Zn-Mg high strength aluminum alloys are well known to have a strong influence not only on the mechanical properties but also on SCC susceptibility. Generally, the increase of matrix precipitate size can reduce the susceptibility to SCC by changing the slip planarity, which also can effectively decrease hydrogen atoms trans-

2 Effects of Microstructure on the Mechanical Properties and Stress Corrosion Cracking of an Al-Zn-Mg-Sc-Zr Alloy 601 ported to the grain boundaries by dislocations to cause embrittlement. 15,17) On the other hand, the larger grain boundary precipitates can trap more atomic hydrogen to nucleate hydrogen bubbles, thereby decreasing the hydrogen concentration at grain boundary below a critical value to prevent intergranular SCC fracture. 26,27) Furthermore, the cathodic grain boundary precipitates grow by depleting solute atoms, broadening the anodic PFZ which contained no strengthen precipitation phase, was soft and weak. 28) The combination of tensile stress and anodic dissolution caused SCC. Most of the previous work discussing the SCC behavior of 7XXX aluminum alloys was based on either two aging treatments, T6 (peakaged) and T7 (overaged) or the improvement SCC susceptibility of the RRA treatment. However, no literature develops specifically for a combination of the same peakaged conditions (T6, RRA and Two-Step) to compare with the mechanical properties and stress corrosion cracking behavior. The purpose of this research is to investigate in detail the effect of microstructure on mechanical properties and stress corrosion cracking behavior of an Al-Zn-Mg-Sc-Zr alloy under the same peakaged conditions (T6, RRA and Two-Step tempers), and to understand the influence of T6, RRA, Two-Step and T7 tempers on the susceptibility of Al- Zn-Mg alloy to SCC. The microstructures of this alloy in each heat treatment condition were investigated by transmission electron microscopy (TEM), with the aim of understanding the relation to SCC susceptibility. For comparison, the electrochemical corrosion of this alloy submitted to different heat treatments was studied by the potentiodynamic polarization tests. 2. Experimental Procedure The composition of the studied alloy determined by inductively-coupled plasma (ICP) spectroscopy was Al- 6.1Zn-2.1Mg-0.12Sc-0.14Zr (mass%), which was supplied in the as-cast condition by Chung-Shan Institute of Science and Technology (Taiwan). 2.1 Heat treatment Ingots were homogenized at 475 C for 24 h in a furnace with an argon protective atmosphere and then cooled in a furnace to room temperature, then hot-rolled at 400 C into plates with a thickness of approximately 2 mm. These specimens were subjected to solution heat treatment at 480 C for 1 h followed by water quenching and then artificial aged to the different conditions in silicone oil bath. Four different tempers have been investigated: T6, T7, RRA and Two-Step. The T6 temper resulted in maximum hardness after aging at 120 C for 48 h, and the T7 temper at 120 C for 96 h. The RRA temper was aged at 120 C for 48 h (T6 condition) and retrogressed at 200 C for 0.5, 1, 2, 4, 5, 8, and 10 min, respectively, water quenched, and then reaged at 120 C for 48 h. The Two-Step temper was first aged at 100 C for 6 h followed by 150 C for 7, 8, 9, or 10 h, respectively. Both RRA and Two-Step tempers resulted in maximum hardness correspond to T6 temper, optimizing the material strength. Fig. 1 The optical micrograph shows grain structure of the extruded rod. ED, ND: extrusion and normal directions, respectively. 2.2 Mechanical and stress corrosion testing The aging response was determined using hardness test by the Vickers apparatus. The tensile properties, including the 0.2% offset yield strength, ultimate strength and elongation, were tested by Shimadzu AG-10TE universal testing machine. These tensile properties were obtained by using tensile specimens have a cylindrical gage section of 3.5 mm in diameter and 14 mm in length with the loading axis parallel to the extrusion direction. A minimum of three specimens, per heat treatment condition, was used for tensile testing. Tensile specimens were tested in air with cross-head speed of 0.2 mm/min. However, the SCC testing procedure was performed according to ASTM G The SCC specimens were prepared from the 15 mm extruded rod and then machined with 3.5 mm diameter and 14 mm gauge length with the tensile axis parallel to the extrusion direction. The loaded (0.75YS) specimens were alternately immersed (10 min immersion, followed by drying in air for 50 min) in a 3.5% NaCl solution (ph between 6.4 and 7.2) for a period of 20 days. A minimum of two specimens, per heat treatment condition, was used for SCC testing. The grain structure of the extruded rod is shown in Fig Potentiodynamic polarization testing In order to compare the results of the SCC, the electrochemical corrosion for all heat treatment conditions have been studied. The specimens were prepared by polishing with 400-, 1000-, 2000-grit papers, then rinsed in acetone and washed in distilled water, and then employed to measure the electrochemical properties by using a potentiodynamic polarization technique. The specimen was connected as the working electrode with a saturated calomel electrode (SCE) and a platinum mesh as the reference and auxiliary electrode, respectively. After the specimen was immersed in the 3.5% NaCl solution at room temperature, and then potentiodynamic scanning at a rate of 0.5 mv/s, from cathodic toward anodic direction, was applied to obtain the polarization curves. 2.4 TEM observation Transmission electron microscope (TEM) was utilized to

3 602 L.-M. Wu, W.-H. Wang, Y.-F. Hsu and S. Trong observe the microstructure of the alloy. Specimens were taken from the hardness- and the SCC-specimens which were ground to a thickness of about 7 mm using SiC paper with a grit from 400 to The thin foils were prepared by double-jet electropolishing at 20 V in a solution of 30% nitric acid and 70% methanol solution cooled to 30 C and observed in a JOEL-100CXII TEM operated at 100 kv. The fracture surfaces of the SCC were observed using scanning electron microscopy (SEM, Philips XL-30). 3. Results and Discussion 3.1 Age hardening Figure 2(a) shows the evolution of hardness curves after aging at 120 C for various times. It can be observed that the hardness value increased initially with the aging time, and then reached a maximum hardness (peakaged T6 temper), finally decreased gradually for the longer aging time. The initially increase is attributed to the nucleation and growth of the large number of G.P. zones and metastable 0 phase. However, the decrease of hardness value for the overaged T7 temper due to the precipitates transformed gradually from the metastable 0 phase to the stable phase. Aging hardening curve for the RRA temper, as shown in Fig. 2(b), illustrate the maximum hardness obtained after RRA temper with retrogression times of 2 min, and the values are close to T6 condition. The Two-Step temper was first aged at 100 C for 6 h followed by 150 C for various times. The change of hardness during the final aging is plotted in Fig. 2(c). It was found that the maximum hardness after final aging for 8 h was reached similar to T6 temper. In this work, it is assumed that the T6, RRA and Two-Step tempers result in the same hardness value, and they were chosen for further investigation on the tensile properties and SCC resistance. All the heat treatment procedures involving the T6, T7, RRA and Two- Step tempers are described in Table Effect of microstructure on the mechanical properties The mechanical properties of the tensile test for various tempered conditions are presented in Table 2. The peakaged conditions exhibited the highest ultimate strength and the smallest elongation. It can be clearly observed from Fig. 3 that the UTS value of RRA temper was similar to the T6 temper, while the Two-Step temper has the lowest value. The differences of strength among these tempers can be explained by the microstructure variation. Figure 4 shows TEM micrographs of the alloy aged to the T6 condition in which the diffraction pattern analysis revealed the presence of 0 phase. The strengthening of the alloy is due to a high density of coherent G. P. zones and nanometer sized semicoherent 0 phase which act as pinning centers for preventing dislocation motion. 29) The T6 temper indicated a peakaged condition which contained mostly semicoherent 0 precipitates and a small amount of coherent G.P. zones in the aluminum matrix made the strength increased. However, the T7 temper indicated an overaged condition and the diffraction pattern analysis revealed the presence of phase (Fig. 5). The larger interparticle spacing and the coarse size of the precipitates in the matrix result in a great decrease in strength. On the Fig. 2 The age hardening curves at the different heat treatments. (a) T6 and T7 tempers; (b) RRA temper; (c) Two-Step temper. Table 1 Heat Treatment Procedures. Temper Designation Aging Treatment T6 (peakaged) 120 C/48 h T7 (overaged) 120 C/96 h RRA (peakaged) T6 aged C/2 min + T6 aged Two-Step (peakaged) 100 C/6 h C/8 h Solution treated at 480 C for 1 h + water quenched

4 Effects of Microstructure on the Mechanical Properties and Stress Corrosion Cracking of an Al-Zn-Mg-Sc-Zr Alloy 603 Table 2 Tensile properties of the alloy for various tempered conditions. Temper Yield strength (MPa) Ultimate tensile strength (MPa) Elongation EL AIR (%) T6 (peakaged) RRA (peakaged) Two-Step (peakaged) T7 (overaged) other hand, the difference in the aging response among the four tempers can be understood by comparing their microstructure, as shown in Fig. 6, present grain boundary precipitates (GBPs) and a precipitated free zone (PFZ) along the grain boundary and strengthening phases inside the grains can also be seen. The PFZ width and the size of grain boundary precipitates were increased from the T6, RRA, Two-Step to T7 temper. Under the same peakaged conditions, the lower strength of the Two-Step temper is primary caused by a wider PFZ and coarser grain boundary precipitates (Fig. 6(c)). Furthermore, the PFZ, in which the solute atoms were depleted and no strengthening precipitation phase existed, was soft and weak. 28) It is further pointed out that the simultaneous reduction in both the PFZ width and the size of the GBPs result in a great increase in strength. Table 2 demonstrated that the RRA temper exhibited better tensile strength, similar to T6 temper, consisting of the fine 0 precipitates in the matrix and with narrow PFZ. In contrast, Fig. 3 Mechanical properties of tensile test for various tempered conditions. Fig. 4 TEM micrographs of the alloy aged under T6 treatment obtained along the [121] Al zone axis. (a) bright-field image; (b) dark-field image. Fig. 5 TEM micrographs of the alloy aged under T7 treatment obtained along the [110] Al zone axis. (a) bright-field image; (b) dark-field image.

5 604 L.-M. Wu, W.-H. Wang, Y.-F. Hsu and S. Trong Fig. 6 TEM micrographs showing the distribution of precipitates and PFZ width of the various tempers. (a) T6 temper; (b) RRA temper; (c) Two-Step temper; (d) T7 temper. Fig. 7 TEM micrographs showing pinning grain boundaries of Al 3 (Sc,Zr) precipitates after at 120 C for 48 h. (a) lower magnification image; (b) higher magnification image. (c) bright field image of Al 3 (Sc,Zr) dispersoids with SAD pattern, [001] zone of Al matrix showing fine superlattice diffraction spots of Al 3 (Sc,Zr) phase(l1 2 structure); (d) dark field image taken from superlattice spot of the fine spherical particles of Al 3 (Sc,Zr) phase.

6 Effects of Microstructure on the Mechanical Properties and Stress Corrosion Cracking of an Al-Zn-Mg-Sc-Zr Alloy 605 Table 3 Results of stress corrosion tensile test for various tempered conditions. Temper Yield strength (MPa) Ultimate tensile strength (MPa) Elongation EL NaCl (%) T6 (peakaged) RRA (peakaged) Two-Step (peakaged) T7 (overaged) Solution: 3.5%NaCl (ph = 6:4 7:2) EL loss (%) Fig. 8 TEM micrographs showing the grain boundary precipitates of the various tempers. (a) T6 temper; (b) RRA temper; (c) Two-Step temper; (d) T7 temper. wide PFZ and coarse grain boundary precipitates of the T7 temper reduce the strength comparing with the T6, RRA and Two-Step tempers (Fig. 6). In addition, the tensile strength of Al-Zn-Mg alloy was increased by additions of Sc and Zr, which was attributed to the pinning of gain boundary by highly stable Al 3 (Sc,Zr) precipitates, as shown in Fig. 7. However, Figure 7(c) shows a SADP of several dispersoids in the grain interior and grain boundary along the [001] zone axis of Al matrix. The pattern revealed the presence of Al 3 (Sc,Zr) phase with L1 2 structure. Fine spherical particles of Al 3 (Sc,Zr) phase can be clearly observed by the dark field image, as shown in Fig. 7(d). 3.3 SCC susceptibility Table 3 presents the mechanical properties of the SCC testing specimens for various tempered conditions. The SCC susceptibility was evaluated from the loss of elongation (EL loss ) defined as follows: EL loss ¼ðEL AIR EL NaCl Þ=EL AIR 100% Increasing EL loss indicated increasing SCC susceptibility. The loss of EI (%) for T7 temper was less than that of the T6, RRA, and Two-Step tempers, indicating that T7 temper was least susceptible to SCC. Therefore, under the same peakaged hardness, the SCC susceptibility was the most severe for T6 temper, intermediate for RRA temper, and minimal for Two- Step temper. The T7 temper possessed the smallest SCC susceptibility. Although the observed difference in behavior of the alloy subject to various heat treatment is not remarkable, it may be explained by difference in microstructure that affects both mechanical properties and stress corrosion carking Microstructural observations Several SCC mechanisms of Al-Zn-Mg alloys have been proposed involving the anodic dissolution, hydrogen-induced cracking and dislocation-assisted cracking ) Typical PFZ widths, grain boundary precipitates and dislocation slip behaviors for various tempers are shown in Fig In the anodic dissolution mechanism, the crack velocity depends mainly on the anodic dissolution rate, i.e. the dissolution rate of grain boundary precipitates. A large size and spacing of grain boundary precipitates will decrease the dissolution rate.

7 606 L.-M. Wu, W.-H. Wang, Y.-F. Hsu and S. Trong Fig. 9 TEM micrographs showing the dislocation slip structure after the SCC tensile testing for the various tempers. (a) T6 temper; (b) RRA temper; (c) Two-Step temper; (d) T7 temper. [121] matrix zone axis and g ¼½111Š for all micrographs. On the other hand, an increasing size of grain boundary precipitates has been proposed to explain the higher SCC resistance in the 7xxx series aluminum alloys. 5) The PFZ width and the size of the particles at the grain boundary of T7 temper, as shown in Fig. 6(d) and Fig. 8(d), are greater than those found under T6, RRA, and Two-Step tempers. The cathodic grain boundary precipitations grow by consuming solute atoms, broadening the anodic PFZ. Reducing the cathode area and increasing the anodic area can decrease galvanic corrosion, reducing susceptibility to SCC. Additionally, the strengthening precipitates in the T7 temper are coarser and less density than those in the T6, RRA, and Two- Step tempers, decreasing the cathodic area, and thereby reducing galvanic corrosion. These results clearly demonstrate that the T7 temper reduces the susceptibility to SCC. In the dislocation-assisted cracking mechanism, several researchers 2,13,16 18) demonstrated that peakaged condition that had low resistance to SCC had long straight narrow slip bands after deformation, while overaged condition that had high resistance to SCC exhibited more homogeneous slip after deformation. A detailed analysis demonstrated that increasing the matrix precipitate can improve the SCC resistance of the alloys, because it can result in the change dislocation slip type from the planar slip to homogenous slip mode. The homogenous slip mode can effectively reduce hydrogen transported to the grain boundary to retard hydrogen embrittlement. 17) In T6 temper, the matrix 0 precipitates and grain boundary precipitates are smaller than a critical size, which exhibits long dislocation lines and planar slip type, as shown in Fig. 9(a). The larger sizes of both the matrix precipitates and grain boundary precipitates in the RRA temper will cause a decrease in the length and density of dislocation lines (Fig. 9(b)), and thereby can improve the SCC resistance of RRA temper. However, the Two-Step temper possesses more homogenous slip mode than the T6 and RRA tempers, as indicated by a comparison of Fig. 9(a) (b) with Fig. 9(c). It is due to a larger average size of matrix precipitates and grain boundary precipitates that can lead to the reduction of planar slip mode. Therefore, the Two-Step temper could effectively improved the SCC resistance comparing with the T6 and RRA tempers. On studying the SCC improvement of Al-Zn-Mg alloy, it is also very important to consider the size of matrix precipitates and grain boundary precipitates. In T7 temper, the sizes of matrix precipitates and grain boundary precipitates are grater than those in the T6, RRA and Two-Step tempers, which make the planar slip more difficult. This results in the decrease of stress concentration at grain boundary and greatly improving the SCC resistance, but the loss of strength. SEM observations of the fracture surfaces of the specimens ruptured in tensile test after SCC testing in the 3.5% NaCl solution for various tempers were shown in Fig. 10. The fractographs of the T6 specimen exhibited some large regions with crack marks and a large amount of intergranular fracture surfaces, as shown in Fig. 10(a) and (b), respectively. In T7 temper, the fractographs exhibited a small amount of short crack marks and some areas with trangranular cleavage-like fracture, as shown in Fig. 10(c) and (d), respectively. For the peakaged tempers, large area fractions of brittle intergranular fracture and some small areas with cleavage-like fracture were present on the T6, RRA and Two-Step tempers. By examinations of tensile fracture surfaces of specimens, the

8 Effects of Microstructure on the Mechanical Properties and Stress Corrosion Cracking of an Al-Zn-Mg-Sc-Zr Alloy 607 Fig. 10 Fracture surfaces of the various tempers, tested in the 3.5% NaCl solution. (a,b) T6 temper; (c,d) T7 temper; (e) RRA temper; (f) Two-Step temper. highest sensitivity of the peakaged tempers can be attributed to intergranular SCC. An overaged temper improves the SCC resistance and reduces the occurrence of intergranular fracture. Hence, the area fraction of intergranular fracture decreased with increasing aging time. 3.4 Electrochemical behavior The polarization curves for different heat treatments are shown in Fig. 11. Adding the Zn and Mg alloying elements to the aluminum solid solution matrix will cause the corrosion potential to become more anodic. 30,31) The concentration of Zn and Mg alloying elements in the matrix will be gradually decreased with increasing the degree of aging due to the precipitation of MgZn 2. The precipitate MgZn 2 is very active and anodic with respect to the film-covered matrix. 31) It is found that by increasing the degree of aging from the T6 to T7 temper, the average size and volume fraction of the anodic 0 and precipitates were increased, but the number of the precipitates were decreased. In other words, the peakaged state had smaller precipitates and a higher precipitate density than that of the overaged state. Therefore, it can be observed that the corrosion potential would become more noble from the T6, RRA, Two-Step to T7 temper (Fig. 11). This means that the T7 temper would possess the best corrosion resistance. The results indicated that the precipitations of the 0 and can reduce the stability of passive film and promote the pitting occurrence. However, the electrochemical corrosion results show the same tendency as the corresponding tests for SCC susceptibility. 3.5 Effects of Sc and Zr on the precipitation microstructure and SCC resistance The addition of Sc to aluminum results in the rapid precipitation of homogeneously distributed Al 3 Sc dispersoids, which are coherent with the matrix and have the L1 2 structure. 32) The presence of Al 3 Sc dispersoids increases the

9 608 L.-M. Wu, W.-H. Wang, Y.-F. Hsu and S. Trong Fig. 11 The polarization curves for the various tempers, tested in the 3.5% NaCl solution. recrystallization resistance of wrought alloys. When both scandium and zirconium are used in the alloy, Al 3 (Sc,Zr) dispersoids form. These dispersoids are more effective recrystallization inhibitors than either Al 3 Sc or Al 3 Zr. 32) In this alloy with scandium and zirconium additions, very fine spherical particles were observed, as shown in Fig. 7. These particles which ranged in size from 20 to 35 nm, helped stabilize the fine-grained microstructure are attributed to the pinning of gain boundaries by fine dispersoid particles. However, a high coherency mismatch is observed for the Al 3 (ScZr) phase (Fig. 7(a)), resulting in significant lattice strain, which serves to block dislocation motion and to prevent grain grow lead to the increase in strength. In addition, it has been reported that an addition of Sc increases corrosion resistance and corrosion potential in Al-Zn-Mg- (Cu) alloys. 32,33) Figure 1 shows the addition of scandium resulted in Al 3 (Sc,Zr) dispersoids that greatly refined grains and restrained recrystallization process. The Al 3 (Sc,Zr) dispersoid particles suppress the planar slip of dislocation, and the smaller grain size restricts the density of dislocation at the grain boundary. Therefore, the concentrations of stress at the grain boundary may be reduced significantly due to the grain refinement effect of Al 3 (Sc,Zr), which decreased the susceptibility to corrosion. The addition of Sc and Zr to high strength Al-Zn-Mg alloys have been found to simultaneously improve the tensile strength and SCC resistance, comparing with the conventional Al-Zn-Mg-(Cu) base alloys. 26,27) 4. Conclusion The effect of microstructure on mechanical properties and stress corrosion cracking of Al-Zn-Mg-Sc-Zr alloy in different tempers was investigated by tensile testing, SCC testing and polarization testing. The following conclusions were drawn: (1) Under the same peakaged hardness, the wide PFZ and coarse grain boundary precipitates of the Two-Step temper lead to the loss of strength comparing with the T6 and RRA tempers. (2) The SCC susceptibility is the most severe for the T6 temper, intermediate for the RRA and Two-Step tempers, and minimal for the T7 temper. The RRA temper can effectively improve the SCC resistance of T6 temper and still possesses better tensile strength. (3) In T7 temper, the sizes of matrix precipitates and grain boundary precipitates are grater than those in the T6, RRA and Two-Step tempers, which make the planar slip more difficult. This results in the decrease of stress concentration at grain boundary and greatly improving the SCC resistance. (4) The fracture characteristics of the peakaged tempers are due to the localized anodic dissolution of grain boundary regions, promoting a damaging intergranular fracture. An overaged temper is failed by a transgranular cleavage-like fracture. (5) The precipitate MgZn 2 is very active and anodic with respect to the film-covered matrix. The peakaged state had smaller precipitates and a higher precipitate density than that of the overaged state. Therefore, it can be observed that the corrosion potential would become more noble from the T6, RRA, Two-Step to T7. (6) The addition of Sc and Zr resulted in Al 3 (Sc,Zr) dispersoids that greatly refined grains and restrained recrystallization process, which enhanced the strength and SCC resistance. Acknowledgements The authors gratefully acknowledge the financial support for this research by Chung-Shan Institute of Science and Technology (Taiwan) under grant no. 95-EC-17-A-08-R REFERENCES 1) M. Puiggali, A. Zielinski, J. M. Olive, E. Renauld, D. Desjardins and M. Cid: Corros. Sci. 40 (1998) ) M. O. Speidel: Metall. Trans. A. 6 (1975) ) M. Reboul: Corrosion sous Contrainte-phenomenologie et mecanismes, Edit. Phys, (1990) ) N. J. H. Holroyd: Proc. Conf. Environm-Induced Cracking, (NACE, Koheler, USA, 1988) pp ) J. K. Parker and A. J. Ardell: Metall. Trans. A. 15 (1984) ) F. Viana, A. M. P. Pinto, H. M. C. Santos and A. Lopes: J. Mater. Process. Technol. 54 (1999) ) J. E. Hatch: Aluminum Properties and Physical Metallurgy, (ASM International, Metals Park, OH, 1984) pp ) D. A. Jones: Principles and Prevention of Corrosion, second ed., (Prentice Hall International Inc., 1977) pp ) B. M. Cina: U.S. Patent , Dec. 24, ) T. D. Burleigh: Corrosion 47 (1991) ) K. Rajan, W. Wallace and J. C. Beddoes: J. Mater. Sci. Technol. 17 (1982) ) I. J. Polmear: Metallurgy of the Light Metals, third ed., (Arnold, London, 1995). 13) T. D. Burleigh: Corrosion. 47 (1991) ) R. H. Jones, D. R. Baer, M. J. Danielson and J. S. Vetrano: Metall. Trans. A. 32 (2001) ) D. Najjar, T. Magnin and T. J. Warner: Mater. Sci. Eng. A. 238 (1997) ) D. Nguyen, A. W. Thompson and I. M. Bernstein: Acta Metall. 35 (1987) ) J. Albrecht, I. M. Bernstein and A. W. Thompson: Metall. Trans. A. 13

10 Effects of Microstructure on the Mechanical Properties and Stress Corrosion Cracking of an Al-Zn-Mg-Sc-Zr Alloy 609 (1982) ) L. Christodoulou and H. M. Flower: Acta Metall. 128 (1980) ) E. N. Pugh: Corrosion. 141 (1985) ) R. G. Song, W. Dietzel, B. J. Zhang, W. J. Liu, M. K. Tseng and A. Atrens: Acta Mater. 52 (2004) ) J. M. Chen, T. S. Sun, R. K. Viswanadham and JAS. Green: Metall. Trans. A. 8 (1977) ) R. K. Viswanadham, T. S. Sun and JAS. Green: Metall. Trans. A. 11 (1980) ) R. G. Song, M. K.Tseng, B. J. Zhang, J. Liu, Z. H. Jin and K. S. Shin: Acta Mater. 44 (1996) ) A. J. Sedriks, J. A. S. Green and D. L. Novak: Metall. Trans. A. 8 (1977) ) J. A. S. Green and W. G. Montague: Corrosion-NACE, 31 (1975) ) T. C. Tsal, J. C. Chang and T. H. Chuang: Metall. Trans. A. 28 (1997) ) T. C. Tsal and T. H. Chuang: Mater. Sci. Eng. A. 225 (1997) ) T. Pardoen, D. Dumont, A. Deschamps and Y. Brechet: J. Mech. Phys. Solids. 51 (2003) ) R. Ayer, J. K. Koo, J. W. Steeds and B. K. Park: Metall. Trans. A. 16 (1985) ) M. Yasuda, F. Weinberg and D. Tromans: J. Electrochem. Soc. 137 (1990) ) S. Aitra and G. C. English: Metall. Trans. A. 12 (1981) ) J. Røyset and N. Ryum: Int. Mater. Rev. 50 (2005) ) Y. L. Wu, F. H. Froes, C. Li and A. Alvarez: Metall. Trans. A. 30 (1999)