(PDF) A study on the microstructure -fracture behavior relations in Al-Si casting alloys

Scripta METALLURGICA Vo1. 30, pP. 47 5- 480, 1994 Pergamon Press Ltd. Et MATERIALIA Printed in the U.S.A. A11 rights reserved A STUDY ON THE MICROSTRUCTURE .FRACTURE BEHAVIOR RELATIONS IN AI.Si CASTING ALLOYS Mahmoud F. Hafiz and Toshiro Kobayashi Toyohashi University of Technology, Production Systems Engineering Deparunent, Japan (Received SePtember 24, 1991) (Revised November L2, 1995) Introduction of the casting alloys, aluminium-silicon (Al-Si) alloys have been widely used in many industries wherc the in addition to excellent castaLility, good corrosion rcsistance as well as low thermal high strength-to-ieighiratio cirrrcienland high ivear resistance (l) are among the properties of engineering interest. Generally, the microstructure ofAl-Si alloys is characterized by two phases i:e Al and Si. Thus, these alloys are a combination of high srrength-brittle phase (Si) and low strength-ductile phase (Al). Therefore, it is beyond doubt that the deformattn behavior and the mechanical properties are controlled by the microstructural featurcs microstnrcture' mechanical properties and of the constituent phases. Developing a better understanding of the is required to satisfy the need of the design fracture behavior relationship of these technically importanialloys the cast products, in order to increase the replacing engineer, on one hand, and ior increasing the confidence in critical stress conditions. Therefore, the present study of more fabricated parts with a cast.o*ion.nt in many concerns itself with the effect of the microstructural parameters in the eutectic region of a hypoeutectic Al-Si The relation between the fracture characteristics casting alloy on the tensile properties and the fracture loughness. well the microstructural parameters, has also been established' and the mechanical prop.r,i'"r,'as as Experimental Procedure The material of concern in the prcsent study is a high purity hypoeutectic Al- 8massToSi alloy. A series of cast produced under a variety of solidification cooling rates with different Sr levels. The ingots of this alloy has been the molten metal into steel, as well as graphite variation in the solidification condition is achieved by casting dimensions with identical cavity. The modified versions of the present alloy moulds having the same shape and treated with different Sr levels, namely 0.017 and 0.03 massTo. Details concerning the melting have been reported elsewhere (2). The modified and nonmodified operation, Sr addition and casting dimensions havebeen uiloyt have been degassed with argon gas prior to casting' : an image analysis system has been To monitor the microstructural features of the alloy under consideration, randomly. At least 5 fields werc analyzed from a single specimen' ,r.airr.'n.rl, oi our"*ations were selected mean value aspect ratio of Si particles The corresponding microstructural parameters considered here are the (AR), the Si particle shape factor (SF), Si-particle equivalent diameter (DE) and the inter-particle spacing to Si- and DE) are defined in Fig. 1' particle diameter, ([ /DE)si. The Si-panicle characteristics (i.e. AR, SF having gauge dimensions of The tensile test is conducted on cylindrical specimens with threaded shoulders and 30 mm in length at a cross-head speed of about 0.5 mm/min. 6 mm in diameter using three p:l-ry bend specimens having a Static fracture toughness characterized by the J-integral is evaluated 15 mm x 15 mm cross-section and span lengtn of 60 mm with a / W - 0.5-0.55, following the multiple specimens technique (3). 475 0956 -7t1x/94 $6. Oo + .00 Copyright (c) 1993 Pergamon Press Ltd' 476 FRACTURE IN A1-Si CASTING Vo1 30, No. 4 AR is the maximum length (Lmax) to the minimum length (Lmin) of the Si-particle (P); DE is the Si-particle equivatent area circle diameter and SF is the area of the circle circumscribing the Si particle (Ac) divided by the equivalent particle area (Ap) Fig. I Definitions of Si-particle characteristics. Fracture parh in the mid-section has been detected using an optical microscope, while the features of the fracrure surface has been delineated using a scanning electron microscope (SEM). Results and Discussions Thble I displays a list of the microstructural parameters obtained for the experimental variables considered here, along with the tensile properties and the fracture toughness (Jq) values. It is obvious that, the major effect of the fast solidificarion cooling rate ( steel mould cast) is used to reduce the size and the distance of eutectic-Si particles, rarher than shape or morphology. This may suggest that the morphology of Si particles is insensitive to the difference in the solidification conditions used in the present investigation.With Sr addition of 0.017 massTo DE, AR and SF are scaled down by a factor of about 10.6, 2.7 and 1.8, respectivelY, for a graphite mould cast and by a factor of about 21.1,2.8 and I .8 for a steel mould cast, compared to those obtained for a nonmodified graphite mould cast alloy. Furthermore, the aspect ratio decreased from 4.1 for the nonmodified graphite mould casr, to 1.3 for the 0.03 massTo Sr-steel mould cast alloy. On the other hand, the circularity of eutectic-Si particles is greatly improved. SF reached a value of about 807o colresponding to a Sr addition of 0.03 massTo and a fast cooling rate and the Si-panicles became much finer. Thus, it can be concluded that, the Sr addition breaks up the silicon parricles in the eutectic phase from a large plate-like structure to fine fibers (2). Thercforc, the eutectic phase of a modified alloys is expected to have a large number of eutectic-Si particles prcsent in its microstructurE. Thus, the decrease in the Si-panicle spacing associated with modification is not surprising. However, the ratio Si- particle spacing to the equivalent particle diameter is higher for the modified alloys. TABLE I The Microstructural Parameters Considered in the Present Investiga- tion Along with the Tensile Properties and Fracture Toughness Values Sr -Content Mould DE AR SF Isi (l/DE)qr YS UTS E JAA (massTo) Tvpe (um) 0tm) (MPa) (MPa) (7o) (kJ/m') 0.000 Graphite 13.10 4.10 2.58 4.80 0.34 58.5 161.0 5.65 44.7 Steel 4.56 3.48 2.41 ' 3.75 0.82 60.4 171.0 7.47 52.4 0.017 Graphite 1.23 1.50 1.44 3.75 3.04 72.5 195.7 14.72 68.8 Steel 0.62 1.46 1.41 2.32 3.76 79.0 216.6 18.84 84.2 3.65 87.0 201.1 21.34 89.4 o.o3o Graphite 3:l? l:13 l:13 1:.ZZ 3.99 94.0 224.3 28.48 92.8 Note that the fracture toughness is reported as Jq, indicating that the valid JIc values arc not obtained. )1. 30, No. 4 FRACTURE IN A1-Si CASTING 477 Tensile Properties Effect of eutectic-Si particle characteristics It can be seen from Table I that while the yield strength (YS) and the ultimate tensile strcngth (UTS) are increased by an appreciable amount due to the change in the Si-particle characteristics, the elongation (E) is drastically increased.The great influence on the ductility can be related to dependency of the response of Si- particles to the deformation of the matrix material in the eutectic region, and cavity formation on the size and morphology of these particles. This is in agreement with results reported previously (4,5) that the local state of stress and the stress acting on the partiele, depend upon the particle size and morphology. Thus, those particles with a large aspect ratio, shape factor and size are expected to nucleate voids at lower applied strainS and stresses. Effect of Si-particle spacine to equivalent Si-pafticle diameter ratio In terms of the ratio of the Si particle spacing to diameter, (I/DE)Si, as shown in Table 1, small (I/DE)Si results in large loss of ductility. Here, it should be mentioned that the small value of ()'/DE)Si is the characteristic of higher shape factor (i.e. poor circularity) and plate-like Si morphology (large aspect ratio), cf. Table 1. Thus, the high stress concentration and large plastic zone in the matrix material around the Si particle could be expected. In this situation, plastic zone overlap in the marix material between two adjacent particles and linking of the voids nucleated at Si particles, will occur at a low strain. In contrast, the increase in (l"7DE)5i concomitant with fibrous Si morphology, shows better SF and smaller AR. Thus, low stress concentration in the matrix material and better load carrying capacity occurs after the void initiation at Si particles lead to failure and after a considerable amount of applied strain. Furthermore, it is interesting to note that the alloys characterized by large (I/DE)Si also have smaller interparticle spacing (cf. Table l).Since the nucleation and growth of voids arc not separable and sequential processes, i.e. new voids are appearing while older ones are growing (6). Thus, a more gradual loss of the load-carrying capacity, after a maximum load with considerable amount of post-maximum load extension, is provided through the smaller particle spacing (7). Fracture Toughness Effect of eutectic -Si particle characteristics Examination of the fracture toughness data complied in Table 1 rcveals that , in general, when the Si-particle tends to be coarser or more slender, the material suffers a loss of its fracture toughness. Increasing the aspect ratio by a factor of about 3.1, SF by a factor of about 2.1 or DE by a factor of about 22.8, reduces the fracture toughness by about 507o. This indicates that the fracture toughness can be related, in some fashion, to the particle characteristics. It has also been shown (8) that increasing the Si-particle size or aspect ratio results in a detrimental effect on the crack initiation strcss. Thus, it can be suggested that low AR and better cinilarity of the Si - particles associated with a Sr modification and fast cooling rate, has affected the first stage of the crack initiation (i.e. void nucleation) owing to an increase in their resistance for fracturc. This explains the dependence of fracnrre toughness of the present alloy on the Si-particles characteristics. Effect of Si - particle spacins to equivalent particle diameter ratio The fracturc toughness of the present alloy is presented in Fig. 2 as a function of (U pE)Si. For the sake of explanation, the fracture toughness as a function of the Si-particle spacing is also imposed on Fig. 2. It is obvious that an incrcase in the ratio (UDE)5i is accompanied by an increase in the fracturc toughness. This improvement in the fracture toughness can be related to the stress-strain state in the matrix material associated with the low aspect ratio and the better circularity of Si panicles, due to Sr addition and fast solidification cooling rate. This in turn, somehow, affects the void $owth rate. On the other hand, if the interparticle spacing has an effect on the fracture toughness of the present alloy, then fracture toughness should increase with increasing interparticle spacing. Hence, this is not the trend in the data presented in Fig. 2, so, one can claim that it is the ratio (UDE)Si 478 FRACTURE IN A1-Si CASTING Vol. 30, No. 4 which has a pronounced effect on the fracture toughness of the present alloy, rather than the interparticle spacing. ;: 100 € ?eo E ''Z 4 a80 '\"':" I cc 100 td z )r., i },Jn E) E E70 i -/l ieo o g ) Eoo & D bso d o.40 .1 i tf i\ ,; ca td z E70 (J p o 80 ,/ 1 IJ t- 50 a 23 E: & or/ l/DE ) sr bs0 6 c t& /|{) t00 200 Ir, ( pm) VGP ( MPe) Fig.2 Fracture toughness of hypoeutectic Al-Si Fig. 3 Variation in the fracturc toughness of hype alloy as a function of Si-particle spacing (1") and eutectic Al-Si alloy as a function of void growth Si-panicle spacing normalized to the equivalent parameter, VGP [ = or*(l/DE) 5il. particle diameter (I / DE)Si. Since the void growth rate seems to be dependent on the (UDE)51, as well as the yield strength of the material, it is relevant to combine these two factors (i.e.(llDE)Si and ) as a void growth "y parameter,VGP[=or*1tr OE) Si]. Figure 3 shows the relation between the fracture toughness and VGP together with the best line fit through the band of the data displayed in this figure. [t is interesting to note that the fracture toughness of the present alloy appears to increase linearly with an increase in VGP. This leads to the suggestion that Hahn and Rosenfield's equation (9) should include a ratio (X.Of) Si rather than 1"g,. Fracture behavior and fracture mechanism Typical fractographs of a coarse and slender eutectic-Si particle (nonmodified, graphite mould cast) and an alloy with a fine fibrous eutectic-Si (steel mould cast, modified with 0.03 massToSr) are shown in Fig. 4. As can be seen in Fig. 4a, the major portion of the fracture surface consists of a complicated array resembling the Si- array in the eutectic region with cloven Si-particles. In addition, broken Si-particles can be detected. The features of Fig. 4a confirm the low fracture toughness displayed by the present alloy in the nonmodified state. Figure 4b, on the other hand, shows two main features, dimple colonies ( much in majority) and a smooth ripple pattern (surrounding the dimple colonies). Figure 4c shows these dimples at a high magnification. These features arc known to be typical ductile fractures. This dramatic change in the fracture pattern can be attributed to the decrcase in the DE and AR in addition to the remarkable improvement in the eutectic particle roundness ( i.e. SF approaching l) associated with the Sr addition ( 0.03 massTo) and fast rate of cooling. The features delineated in Fig. 4b, explain the significant improvement in the fracture toughness,on one hand, and reflect the importance of Si-particle size and morphology, on the the other hand. Vol. 30, No. 4 FRACTURE IN A1-Si CASTING 479 Fig. 4 The fracture surfaces of the present alloy (a) nonmodified graphite mould cast (i.e.coarse, plate-like eutectic - Si) , (b) steel mould cast, modified with 0.03 massToSr ( i.e.fine-fibrous eutectic-Si morphology) and (c) as in (b) but at a high magnification. Optical fractographs of the mid- section are shown in Fig.5. It is found, in general, that voids are initiated at Si- particles. The individual voids grow and coalesce, creating microcracks in the eutectic region. Then, these micre cracks would link up to the main crack (Fig. 5a). It is worthy to mention that the change in the Si-particle characteristics does not have an influence on the sequence of fracture process, but it indeed affects the amount of the work done by the material to proceed these stages of fracture. In such cases, the tensile data can provide an estimation of the energy spent in void nucleation, growth and coalescence. The area under the engineering stress- strain curve, prior to and after the ultimate stress is given in Table 2. It is apparent that the energy spent in void nucleation, growth and coalescence is significantly affected by the change in the eutectic-Si particle characteristics presented in Table l, owing to Sr addition and solidification cooling rate. This result is supportive of the inhancement of the fracture toughness of the present alloy associated with the change in eutectic-Si. It can further be seen from Fig. 5a that the crack prefers to propagate through the eutectic rcgion, circumventing the morc tough phase (i.e Al-dendrites), whenever it is possible. However, in a modified alloy, Fig. 5b, the crack tends to propagate transgranularly through the Al-dendrites. This can be attributed to the more uniform distribution of the moit important phase (i.e. eutectic mixture), owing to the decrease in the dendrite spacing associated with the Sr modification and fast cooling rate (2,10). Fig. 5 Optical micrographs show the fracture path in the mid-section of the present alloy (stccl mould cast), (a) nonmodified and (b) modified with 0.03 mass%Sr. 480 FRACTURE IN A1-Si CASTING Vo1 30, No. 4 TABLE 2 Work per Unit Yolume Done on the Material Prior to Rupture (estimated from the area under engineering stress-strain curve). Sr - Content Mould Pre-maximum Stress Post-maximum Stress - (massTo) Type Work Done (MJ / m31 Work Done (MJ / mr) Graphite 4.76s L.434 0.000 6.185 2.460 Steel 0.017 Graphite 17.0L7 2.727 Steel 22.533 5.302 Graphite 0.030 20.639 10.104 Steel 27.738 17.579 Conclusions The effecs of the microstrucnrral feattres in the eutectic rcgion on tho tensile propertics, fracture toughness and fracture behavior of high purity hypoeutectic Al-Si casting alloy have been investigated. Based on the results obtained in the present study, the followilpg conclusions can be drawn : l. The combined effect of the Sr modification and fast solidification cooling rate has a pronounced influence on the microstructural features of the eutectic region. Si-particle size, aspect ratio and shape factor are scaled down by a factor of about 22.8, 3.1 and 2.1, respectively, owing to a Sr addition of 0.03 massTo (steel mould cast). 2. The break up of thc coarse-platelike Si-particles into fine-fibrous ones decreased the interparticle spacing. However, the interparticle spacing normalized to the particle diameter t(VOel t,1 is increased with the dramatic decrease in particle sizc. 3. Si-particles with a fine, low aspect ratio and better circularity resulted in a significant improvement in the tensile properties and the fracture toughness of the present alloy owing to their influence on the void nucleation process. On the other hand, increasing the value of [ (VDE) 5;] has affected the mechanical properties in two ways, decreasing the void growth rate and providing higher load carrying capacity. 4. Fracturc toughness data is found to correlate well with VGP [= oy * (I/DE) 5;J. This suggests that Hahn and RosenFreld's equation (9) should include a ratio (UOe15; rather than 1.51. 5. Independent of the mrcrostructural features in the eutectic region, the fracture mechanism of the present alloy follows the sequence; void nucleation at Si-particles, subsequent growth and coalescence of the individual voids. However, the fracturc pattern showed geat dependency on the microstructure of the eutectic phase. References l. L.F. Mondolfo, " Aluminium Alloys; Structure and Properties", Butterworth, London, (1976). 2. M.F.Hafrz and T.Kobayashi, Trans. Japan Foundrymen's Soc. (In Press). 3. ASTM, E8l3-83, (1983). 4. K. Wallin et al. , Int. I. of Fracturc,32 201 (1987). 5. A. Saigal, AFS Trans. 94,219 (1986). 6. L.E. Roy et al, Acta Metall., 29,1509 (1981). 7. P.E. Magnusen, E.M. Dubensky, and D.A. Koss, Acta Metall. 36 (6), 1503 (1988) . 8. A. Saigal and J.T. i3erry, AFS Trans.,93, 699 (1985). 9. G. T. Hahn and A..R.. Rosenfield, Metall. Trans., 6A, 653 (1975) . 10. M.F. Hafiz and T. )iiobayashi, Proc. of the Int. Conf. on the Mech. behavior of Ductile Cast Iron and other Cast Metals, Kitakyushu, Japan, July 31- Aug.l (1993), In Press.