- Email: [email protected]
The quasi-static, cyclic fatigue and final fracture behavior of a magnesium alloy metal-matrix composite
Composites: Part B 36 (2005) 209–222 www.elsevier.com/locate/compositesb
The quasi-static, cyclic fatigue and final fracture behavior of a magnesium alloy metal-matrix composite T.S. Srivatsan*, Meslet Al-Hajri, P.C. Lam Division of Materials Science and Engineering, Department of Mechanical Engineering, College of Engineering, University of Akron, Akron, OH 44325-3903, USA Received 10 April 2004; revised 23 September 2004; accepted 23 September 2004 Available online 23 November 2004
Abstract In this paper the results of a study to understand the role of composite microstructure on quasi-static, cyclic fatigue, deformation and fracture behavior of magnesium alloy Z6 discontinuously reinforced with silicon carbide (SiC) particulates are presented. Quasi-static fracture of the composite was dominated by cracking of the reinforcing particulates present in the magnesium alloy metal matrix. Final fracture occurred as a result of crack propagation through the alloy matrix between particulate clusters. The composite specimens were cyclically deformed over a range of stress amplitudes. The stress amplitude–fatigue life response was found to degrade with an increase in test temperature. Influence of test temperature on quasi-static and cyclic response and final fracture behavior of the composite is discussed in light of the concurrent and mutually interactive influences of composite microstructural effects, deformation characteristics of constituents of the composite material, magnitude of cyclic stress amplitude and fatigue life. q 2004 Elsevier Ltd. All rights reserved. Keywords: Magnesium alloy; Particulate reinforcement; Composite; Microstructure; Fatigue; Fracture; Temperature
1. Introduction A growing and sustaining interest arising primarily from the need to improve fuel economy, minimize vehicle emissions, enhance styling options, improve overall performance, while concurrently maintaining safety, quality, reliability, durability while also ensuring profitability, are some of the challenges that most certainly need to be addressed by emerging and future generation materials that find use in a variety of performance-critical applications in the industries spanning aerospace, automotive and ground transportation. This burgeoning need has provided the impetus for the development, emergence and rapidly increasing use of a ‘new’ class of metallic materials, namely, metal-matrix composites (MMCs) [1–4]. Their exceptionally high specific modulus (E/r), strength-toweight (s/r) ratios, fatigue strength, wear resistance, and tailorable properties such as thermal expansion coefficient,
* Corresponding author. Tel.: C1 330 972 6196; fax: C1 330 972 6027. E-mail address: [email protected] (T.S. Srivatsan). 1359-8368/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.compositesb.2004.09.004
are some of the important attributes that have helped make MMCs attractive and potentially viable candidates for selection and use in several weight-sensitive, stiffnesscritical and performance-worthy components in both air and ground transportation systems [2–8]. Driven initially by the rising demand for high performance military and space applications, the discontinuously reinforced metal matrices have most certainly stimulated considerable scientific and technological interest in the last two decades. Rapid strides in recent years have seen their use for a wide range of applications, specifically recreational [2,4,8], automotive [9] and some aerospace products [3,4]. Also, the emerging need for satisfying high fuel-economy goals in the automobile industry will be a major challenge for which the metal-matrix composites based on the lightweight alloys will play a significant role. However, problems arising from inadequate fracture toughness, limited damage tolerance, poor tensile ductility, inferior fracture-related properties, size limitations on available product forms, and intrinsic material variability, continue to pose a severe limitation to the wide spread application and use of the metal-based composites [2,4,5,10].
210
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
A class of metal-matrix composite materials are attracting use in several automotive components and consumerrelated applications, is the magnesium alloy matrices discontinuously reinforced with particulate, whisker or short-fiber reinforcements. These discontinuously reinforced metal matrices will be referred to henceforth in this paper as DRMg MMCs. The presence of a discontinuous reinforcement phase in a continuous magnesium alloy metal matrix can result in properties not attainable by other means, thus extending the possible applications [7]. Both die cast and extruded Mg/SiC particle composites exhibit attractive physical properties. In fact die cast bars of AZ91/ SiC have shown an increase in elastic modulus, offset yield strength and ultimate tensile strength as the percentage of SiC reinforcement increases [10]. At room temperature the metal matrix reinforced with 25% SiC by volume had an elastic modulus 72% higher than that of the unreinforced alloy. The yield strength was 47% greater and the ultimate tensile strength was 23% greater than the unreinforced counterpart. With an increase in temperature, the properties deteriorated, but in the same manner for both the reinforced metal matrix and the unreinforced counterpart. An ability to achieve improvements in mechanical properties is largely dependent on the mutually interactive influences of: (a) intrinsic properties of the composite constituents, i.e. metal matrix and reinforcement, and (b) the size, shape, orientation, volume fraction and distribution of the reinforcement phase in the metal matrix [11–14]. Selection of type of reinforcement and its geometry (shape) is critical to obtain the optimum combination of properties at a substantially low cost [15]. Of the available choices silicon carbide particulates (SiCp) are the preferred reinforcements primarily because enhanced properties can be achieved with little or no penalty in density [9]. From an economic standpoint, the use of ceramic particulates as the reinforcement phase is advantageous. The SiC particulate reinforcement is easily available and relatively inexpensive. The magnesium alloy metal matrix reinforced with ceramic particles is a promising material based on its potential for low cost [16]. From a design engineering perspective, the attractiveness in preferring and choosing a discontinuously reinforced magnesium alloy MMC for many applications stems from an improvement in specific modulus, that is, the density compensated increase in elastic modulus. The discontinuously reinforced magnesium alloy metal matrices (DRMg MMCs) can be tailored so as to have properties that are near isotropic. Furthermore, discontinuously reinforced magnesium alloy metal matrices maintain their amenability to: (a) conventional metallurgical processing, (b) fabrication, and (c) characterization methods currently used for the unreinforced counterparts [2,4,5,10]. Several practical applications for the DRMg MMCs involve cyclic loading, and therefore knowing the fatigue response coupled with an understanding of the resultant fracture characteristics is essential. Reinforcing a magnesium alloy metal matrix can have a beneficial or
detrimental influence on cyclic fatigue resistance depending upon diverse and yet related factors such as: (i) method of synthesis (primary processing method), (ii) reinforcement type, (iii) geometrical constitution (size, shape, volume fraction, and distribution of the reinforcing phase), (iv) nature of secondary processing (aging treatments), and (v) strength of the metal matrix and reinforcement/matrix interfaces. The presence of discontinuous particulate reinforcements in a magnesium alloy metal matrix does exert an influence on crack growth behavior, i.e. extent and severity of damage propagation, and resultant fracture characteristics. The influence on crack growth, or damage propagation, is governed by the mutually interactive influences of matrix composition and microstructure, geometrical constitution, processing conditions and applied cyclic stress intensity range. While few studies have focussed on developing an understanding of the nature and influence of the reinforcement particles on matrix microstructure and tensile response, the fatigue database for discontinuously reinforced magnesium alloy matrices is limited, and there certainly exists a need to expand this aspect of mechanical characterization. Presence of discontinuous particulate reinforcements in a ductile magnesium alloy metal matrix will alter precipitation kinetics of the material during heat treatment (i.e. aging treatment) and thus the microstructure of the composite compared with the unreinforced monolithic counterpart [13,17]. A change in intrinsic microstructural features will exert an appreciable influence on mechanical response and resultant fracture behavior. The objective of this paper is to understand the influence of discontinuous particulate reinforcement on tensile, cyclic fatigue and fracture (quasi-static and cyclic) behavior of a magnesium–zinc alloy reinforced with fine particulates of silicon carbide (henceforth referred to as SiCp). The tensile deformation, fatigue response and fracture behavior of the composite are discussed in light of the concurrent and mutually interactive influences of composite microstructural effects, matrix deformation characteristics, test temperature and nature of loading.
2. Material Elemental magnesium was melted in a stainless steel crucible. The melt was prevented from burning by using a protective atmosphere of carbon dioxide, air and SF6 [18]. A nominal 6.0 wt% of zinc was added to the melt. This was followed by the addition of silicon carbide (SiC) particulates (mesh size: 600 grit) to the molten magnesium–zinc alloy using proprietary techniques. The molten composite material was cast into billets. During casting a stainless steel impeller was used to continuously stir the composite melt so as to retain the SiC particulates in suspension. Prior to extrusion, the composite billet was homogenized at 672 K for 4 h followed by quenching in cold water. The billet was
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
extruded into a solid rod using extrusion techniques similar to that used for conventional magnesium. The reduction ratio used was 60:1. The manufacturer (DOW Chemical Company at Freeport, TX, USA) considers precise details of the processing treatments (primary and secondary) as proprietary. The Z6/SiC/20p composite material was provided in the as-extruded condition (denoted henceforth by the symbol ‘F’).
3. Experimental techniques 3.1. Specimen preparation Blanks of length 150 mm were cut from the as-received Z6/SiC/20p-F composite extrusion using a diamond-coated blade. Smooth cylindrical test specimens (gage lengthZ 6.25 mm and gage diameterZ25 mm) were precision machined from the blanks using diamond tooling. The specimens were machined with the stress axis parallel to the extrusion direction and conformed to specifications outlined in ASTM Standards E8-93 [19] and E606-93 [20]. The length-to-diameter ratio of the mechanical test specimen was chosen so as to minimize buckling during fully reversed (tension–compression) stress amplitude-controlled cyclic deformation. To minimize the effects of surface irregularities and finish, the test specimen surface was prepared by mechanically polishing the gage section using progressively finer grades of silicon carbide impregnated emery paper and then finish polished using 0.5 mm alumina powder suspended in distilled water so as to obtain a mirror like finish that is free of all circumferential scratches and surface machining marks. 3.2. Initial microstructure characterization Metallographic samples were cut from the as-received composite extrusion, mounted in bakelite and wet ground on 320, 400 and 600 grit SiC impregnated emery paper, using copious amounts of water as lubricant. The mounted and ground samples were then mechanically polished using a 1 mm alumina-powder suspended in distilled water. Fine polishing to near mirror like finish was achieved using 0.5 mm diamond paste. Reinforcement morphology and its distribution in the metal matrix along with other intrinsic microstructural features were identified by examining the samples in an optical microscope and photographed using bright-field illumination technique.
211
The tests were conducted in controlled laboratory air environment (relative humidity of 55%) at ambient temperature (27 8C). Tensile tests were conducted in accordance with procedures outlined in ASTM E8 at a strain rate of 1!10K4 sK1 and using a 12.5 mm gage length extensometer to measure strain. Reduction in area was measured at the fracture surface of the failed specimen using a micrometer. In considering the limited tensile ductility, cyclic stress resistance, and high-cycle fatigue (HCF) life of the silicon carbide particulate reinforced magnesium alloy at room (ambient) temperature, the stress amplitude-controlled highcycle fatigue tests were also conducted at an elevated temperature. The elevated temperature chosen was 150 8C and corresponded to the maximum temperature limit of the environmental chamber unit. The elevated temperature tests were conducted inside an environmental chamber unit (Instron: Model 3111). The temperature was controlled with the aid of a temperature controller in conjunction with a thermocouple. The thermocouple was fixed onto the specimen’s surface. Maximum temperature variation was well within 2 8C of the set-point temperature (150 8C) over the entire duration of the test. Before each test, the specimen was maintained at the test temperature for 30 min so as to achieve stability with the environment. The fatigue tests were conducted at a constant cyclic frequency of 5 Hz and at a stress ratio (RZsminimum/smaximum) of 0.1. The number of cycles to cause complete failure or separation is taken as fatigue life (Nf). 3.4. Failure-damage analysis Fracture surfaces of the tensile and cyclically deformed test samples were examined in a scanning electron microscope (SEM) to: (a) determine the macroscopic fracture mode, and (b) characterize the fine-scale topography to establish the microscopic mechanisms governing fracture. The distinction between macroscopic mode and microscopic fracture mechanisms is based on the magnification level at which the observations are made. The macroscopic mode refers to the nature of failure, while the microscopic mechanism relates to the local failure processes (microscopic void formation, coalescence, and nature and intensity of cracking). Samples for SEM observation were obtained from the failed specimens by sectioning parallel to the fracture surface. Matching fracture surfaces were viewed to determine the presence of fractured SiCp on both halves of the specimen.
3.3. Mechanical testing 4. Results and discussion Test specimens were precision machined in accordance with procedures outlined in standards ASTM E8 and E606. The tensile and cyclic fatigue tests were performed on a fully automated, closed-loop servohydraulic test machine (INSTRON) equipped with a 10,000 kg (98 kN) load cell.
4.1. Undeformed microstructure Fig. 1 is the optical micrograph showing microstructure of the Z6/SiC/20p composite The silicon carbide
212
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
Fig. 1. Optical micrograph illustrating microstructure of the Z6/SiC/20p-F composite.
particulates (SiCp), in the Z6 magnesium alloy metal matrix, were generally of near-uniform size. However, a few irregularly shaped particulates were found dispersed randomly between the near uniform shaped particulates. An agglomeration or clustering of the SiCp was observed. An agglomerated site consisted of the larger SiC particulates intermingled with the smaller, uniform and more regularly shaped SiC particles. Alignment of the reinforcing SiC particulates was observed in the direction of billet extrusion. Such an alignment has been documented in other extruded aluminum alloy-based MMCs [21–24]. However, in this study no attempt was made to determine the particle size distribution for the Z6/SiC/20p-F MMC material. The etched composite microstructure revealed the matrix to consist of very fine grains, which could not be easily resolved in the optical microscope at magnifications !1000!. 4.2. Tensile properties and fracture The tensile properties of the Z6/SiC/20p-F composite, at the two test temperatures, are summarized in Table 1. The results are the mean values based on duplicate tests. At ambient temperature (27 8C), the composite material revealed limited ductility, quantified in terms of both
elongation-to-failure (3f) and reduction-in-area (RA). An increase in test temperature resulted in a 46% decrease of yield strength (sys) and a 50% drop in ultimate tensile strength (sUTS) but a concomitant improvement in: (a) elongation-to-failure (3f) by over 800%, and (b) the reduction in area by 500%. It is evident that the degradation in strength had a profound influence on tensile ductility of this composite. The tensile fracture surfaces provide valuable information pertaining to microstructural effects on tensile ductility and fracture properties of the magnesium metal matrix composite. It is fairly well documented that the fracture of unreinforced monolithic alloys occurs by events of microscopic void nucleation and growth, with the nucleation essentially occurring at the coarse intermetallic particles and other insoluble phases in the microstructure [25,26]. An essential requirement for void nucleation either at a reinforcing particle and/or other coarse second-phase particles in the microstructure is the development of a critical normal stress across the particulate-matrix interfaces [27]. In the unreinforced metal matrix, the nucleation of cavities and voids is favored by the concurrent and mutually interactive influences of: (1) cracking of the hard, intrinsically brittle and elastically deforming SiCp inclusions, and (2) decohesion at interfaces between the hard and brittle reinforcing SiC particulate and the soft and ductile magnesium alloy metal matrix. The Z6/SiC/20p-F composite exhibited limited ductility at ambient temperature, with fracture, on a macroscopic scale, occurring on a plane normal to the far-field stress axis. High magnification observations of the tensile fracture surface revealed microscopic features reminiscent of locally ductile and brittle mechanisms. Representative fractographs of the tensile fracture surface at the two temperatures are shown in Figs. 2–4 and discussed in the following sections. 4.2.1. Test temperature of 27 8C Tensile fracture of the Z6/SiC/20p-F MMC at room temperature was flat, near featureless and essentially on a plane normal to the far field stress axis when viewed on a macroscopic scale (Fig. 2a). Microscopic observations revealed cracking parallel to the major stress axis (Fig. 2b). The macroscopic cracks were surrounded by cracked particulates and ductile regions, which are referred to as tear ridges. Isolated tear ridges were evident on the tensile fracture surface (Fig. 2c). The matrix was covered with numerous microscopic cracks (Fig. 2d) and cracked SiC particulates (Fig. 3a). The overload region revealed
Table 1 Tensile properties of Mg-6Zn/SiC/20p-F composite Temperature (8C)
Elastic modulus (GPa)
27 150
66 48
Yield strength
Ultimate tensile strength
MPa
ksi
MPa
ksi
307 166
44.5 24
379 186
55 27
YS/UTS (%)
Elongation (%)
Reduction in area (%)
ln(A0/Af) (%)
81 89
2.0 19.0
3.0 19.0
3.2 21.0
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
213
Fig. 2. Scanning electron micrographs of the tensile fracture surface of Z6/SiC/20p-F composite at ambient temperature (27 8C).
Fig. 3. Scanning electron micrograph of the tensile fracture surface of the Z6/SiCp composite showing: (a) cracked SiCp, and (b) fine microvoids and isolated shallow dimples.
214
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
a population of microscopic voids of varying size and shapes intermingled with ductile dimples (Fig. 3b), features reminiscent of locally ductile fracture. 4.2.2. Test temperature of 150 8C On a macroscopic scale, the tensile fracture surface of the Z6/SiC/20p-F MMC deformed at the elevated temperature (150 8C) was flat and similar to the ambient temperature fracture surface but relatively rough when viewed on a microscopic scale (Fig. 4a). Isolated pockets of macroscopic cracks were distributed randomly through the fracture surface (Fig. 4b). The fracture surfaces revealed the presence of cracked SiC particulates surrounded by a population of shallow dimples. The matrix of the composite was covered with a population of microscopic voids of a wide range of sizes (Fig. 4c). The voids were distributed throughout the fracture surface and were intermingled with shallow dimples (Fig. 4d). The constraints in mechanical deformation imposed by the hard, intrinsically brittle and elastically deforming SiCp
in a relatively soft and ductile magnesium alloy metal matrix coupled with the concurrent development of a triaxial stress state in the matrix of the Z6/SiCp composite is a factor that strongly aids in limiting flow stress of the composite microstructure, while concurrently placing constraints on microscopic void nucleation and growth. As a direct consequence of the deformation constraints induced by the discontinuous SiC particulates, a higher applied stress is required to initiate plastic deformation in the microstructure of the metal-based composite. This translates to a higher elastic stiffness and yield strength of the Z6/SiC/20p-F composite when compared to the unreinforced Z6 alloy counterpart. Failure of the reinforcing SiC particulates either by cracking or decohesion at their interfaces with the magnesium alloy metal matrix can be attributed to the mutually interactive influences of the following: (a) local plastic constraints, (b) particle size, and (c) degree of particle agglomeration. The local plastic constraints are particularly important for the larger-sized SiC particles and particle
Fig. 4. Scanning electron micrographs of the tensile fracture surface of Z6/SiC/20p-F composite at 150 8C.
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
clusters during composite fracture [28–30]. Examination of the tensile fracture surfaces revealed the damage associated with fracture to be localized at the discontinuous SiCp reinforcements with little evidence of void formation away from the fractured SiC particulates. Fracture of the hard, brittle and elastically deforming SiCp was found to be greater in regions of particle clustering. This is attributed to: (1) enhanced local stresses resulting from restriction of plastic deformation, and (2) an intrinsic brittleness of the reinforcing SiCp with the propensity for it to fracture due to localized deformation. The two concurrent and mutually competitive factors result in: (a) particulate failure by cracking, and (b) aided by decohesion or separation at its interfaces with the metal matrix, being the dominant damage modes. The damage of the composite microstructure, during uniaxial loading, arising from the synergistic influences of particle cracking and decohesion at the magnesium alloy matrix–SiC particulate interfaces results in a detrimental influence on tensile ductility. Furthermore, assuming that the metal matrix-reinforcement particle interfaces are strong, the triaxial stresses generated during far-field tensile loading favor limited growth of the microscopic voids in the matrix of the composite. Under the influence of an applied tensile load, the microscopic voids appeared to have undergone limited growth confirming a possible contribution from SiC particle constraint-induced triaxiality on deformation and failure of the composite matrix. The limited growth of the microscopic voids without their coalescence, as a dominant fracture mode for the Z6/SiC/20p-F6 composite, suggests that the deformation properties of the magnesium alloy metal matrix are significantly altered by the presence of discontinuous SiCp reinforcements. Fracture of the brittle SiC particulates aided by failure, or decohesion at the matrix–particulate interfaces results in the formation of voids. Few of the fine microscopic voids coalesce and the halves of these voids are the shallow dimples observed on the tensile fracture surface (Fig. 3b). The lack of formation of dimples, as a dominant fracture mode, is attributed primarily to the constraints in plastic flow of the composite matrix caused by the presence of the discontinuous SiCp reinforcement phase and not due to limited ductility of the Z6 magnesium alloy matrix. The constraint on plastic flow favors the formation of fine tear ridges between the SiC particulates in the magnesium alloy metal matrix.
215
Fig. 5. Influence of temperature on cyclic stress amplitude (Ds/2)-fatigue life (Nf) response of the Z6/SiC/20p-F composite.
essentially unaffected during stress controlled cyclic deformation. The results of the axial stress amplitude controlled tests are shown in Fig. 5 as the variation of stressamplitude (Ds/2) with fatigue life (cycles to failure NF). The best curve was drawn through the test data points. Coincidentally, the stress amplitude (Ds/2)–fatigue life (Nf) curve exhibits the general trend shown by the non-ferrous metals, i.e. increasing fatigue life with decreasing stress amplitude levels. At equivalent stress amplitudes an increase in test temperature resulted in degradation in cyclic fatigue life by at least an order of magnitude. To understand and establish the reasons responsible for the observed degradation in life, the stress amplitude–fatigue life data are replotted in terms of elastic strain amplitude in the manner suggested by Hassen and co-workers [30], where at a given test temperature the elastic strain amplitude (D3e/2) is the stress amplitude (Ds/2) at that temperature normalized by elastic modulus (E) at that temperature. Variation of elastic strain amplitude with fatigue life is shown in Fig. 6. This accounts for the decrease in stiffness
4.3. High cycle fatigue behavior The stress amplitude-controlled high-cycle fatigue test provides information pertinent to the existence of an endurance limit. In the low stress amplitude and resultant high cyclic fatigue life of Mg MMC the property changes are expected to occur at and around the vicinity of the fracture process zone, i.e. the small region immediately ahead of an advancing crack tip where crack extension originates, while the material in the far field remains
Fig. 6. Elastic strain amplitude (D3e/2)–fatigue life (Nf) response of the Z6/SiC/20p-F composite.
216
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
of the Z6/SiC/20p-F composite material with an increase in test temperature. A minimal difference in elastic strain amplitude versus fatigue life behavior is observed at the two test temperatures suggesting that the demonstrated degradation in cyclic fatigue life, with an increase in test temperature, is a reflection of the lower elastic modulus (48 GPa) of the composite material at the higher test temperature (150 8C) when compared with the modulus (66 GPa) at ambient temperature (27 8C). To further examine the influence of temperature on stress amplitude controlled fatigue response, the stress–fatigue life data are replotted to account for differences in strength of the composite material at the two temperatures. For purposes of rationalizing influence of composite material strength (yield strength versus ultimate tensile strength) on stress amplitude versus fatigue life response the stress amplitude was normalized with strength at that temperature so as to evaluate the feasibility of using the composite in performance-critical applications. In Fig. 7A the stress amplitude (Ds/2) has been normalized by the uniaxial yield strength
Fig. 7. (A) Variation of cyclic stress amplitude normalized with yield strength (sYS) versus fatigue life (Nf) for the Z6/SiC/20p-F composite at the two temperatures. (B) Variation of cyclic stress amplitude normalized with ultimate tensile strength (sUTS) versus fatigue life (Nf) for the Z6/SiC/20p-F composite at the two temperatures.
(sYS) of the composite at the temperature, while in Fig. 7B the cyclic stress amplitude (Ds/2) has been normalized by the ultimate tensile strength. (sUTS) of the composite at the temperature. In both cases the normalized stresses are plotted against the number of cycles-to-failure (Nf). In both situations the Z6/SiC/20p-F composite material reveals improved high cycle fatigue life at the higher test temperature. The improvement is significant at the higher ratios of stresses and resultant low fatigue life than at the lower ratios of stresses and resultant enhanced fatigue life. The beneficial effects of elevated temperature on cyclic fatigue resistance at the higher ratios of stress are a direct indication of the superior crack initiation and/or crack growth resistance of the composite material at the elevated temperature. 4.4. Cyclic fracture behavior Examination of the fracture surfaces of the cyclically deformed fatigue specimens, in a JEOL scanning electron microscope, was done at: (a) low magnification to identify the regions of fatigue initiation and final fracture (overload), and (b) at higher magnifications in the fatigue region to identify the regions of microcrack formation and early microscopic crack growth, and the overload region to identify the fine-scale fracture features. Representative fracture features of the Z6/SiC/20p-F composite are shown in Figs. 8–11. 4.4.1. Ambient temperature (27 8C) High cycle fatigue fracture of the composite sample deformed at the maximum cyclic stress of 283 MPa (NfZ6369 cycles) revealed fracture to have essentially occurred on a plane that is normal to the far-field stress axis. The fracture surface revealed a small region of fatigue, about the size of a thumbnail, and a large portion to be overload failure (Fig. 8a) with little evidence of gross cyclic ductility. Low magnification observations revealed randomly distributed microscopic cracks in the region of early crack growth (Fig. 8b); features reminiscent of locally brittle failure. The microscopic cracks were common in test specimens deformed at the higher cyclic stress and resultant short fatigue life. There was little evidence of crack branching suggesting that the damaged region ahead of the crack tip be greatly reduced. High magnification observations of the overload region revealed a population of fine microscopic voids of varying size and shapes and shallow dimples (Fig. 8c) and cracked SiC particulates (Fig. 8d), features reminiscent of locally ductile and brittle failure. The fine microscopic voids were distributed through the fracture region with the occurrence of microvoid coalescence. The composite sample cyclically deformed at the lower stress of 184 MPa with resultant enhanced fatigue life (NfZ 162,197 cycles) revealed distinct regions of stable crack growth and overload (Fig. 9a). High magnification
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
217
Fig. 8. Scanning electron micrographs showing HCF fracture surface features of the Z6/SiC/20p-F composite deformed at cyclic stress amplitude of 283 MPa, NfZ6369 cycles, showing: (a) overall morphology, (b) array of microscopic cracks, (c) voids of varying size and shallow dimples on the overload fracture surface, (d) cracked of SiC particulate.
observations in the region of stable crack growth revealed it to be flat with an array of fine microscopic cracks (Fig. 9b). The overload region revealed microscopic cracks, isolated pockets of shallow dimples and microscopic voids (Fig. 9c), features reminiscent of locally brittle and ductile failure. The randomly distributed voids revealed evidence of microvoid coalescence (Fig. 9d). The microscopic voids ranged in size from the sub micron to tens of microns. 4.4.2. Elevated temperature (150 8C) At this temperature the fatigue fracture surface of the composite sample cyclically deformed at the lower stress amplitude and resultant longer fatigue life (Fig. 11) was quite similar to that observed at the higher stress amplitude and short fatigue life (Fig. 10). In addition to cracking of the reinforcing particulates, decohesion or separation, promoted by local stress concentration arising from dislocation buildup during cyclic deformation, was evident at the metal
matrix–particulate interfaces. The larger SiC particles fracture at a much lower stress than the smaller SiC particles. This is because the intrinsic flaw size in a reinforcing SiC particle is directly proportion to its size. Assuming flaw size (say ‘d’) is a fraction (say ‘x’) of the diameter of the particle (2r), the stress essential to fracture or rupture the second-phase/reinforcing particle is sfracture Z Kpfracture =ð2pDÞ0:5 p
(1)
where Kpfracture is the fracture toughness of the reinforcing SiC particle. Further, the larger dimensions of the reinforcing SiCp facilitate a greater degree of stress transfer from the plastically deforming metal matrix to the elastically deforming particulate. The smaller SiCp do not crack easily. However, on account of intrinsic strain differences between the plastically deforming matrix and the elastically deforming particle, the process culminates in separation, or decohesion, at its interfaces with the metal
218
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
Fig. 9. Scanning electron micrographs showing HCF fracture surface features of the Z6/SiC/20p-F composite, deformed at cyclic stress amplitude of 184 MPa, NfZ162,197 cycles, showing: (a) overall morphology, (b) high magnification of (a) showing randomly distributed microscopic cracks on transgranular fracture surface, (c) microcrack, dimples and voids, (d) void coalescence.
matrix. The tendency for SiC particulate fracture is higher in regions of high local particulate volume fraction. This arises from increased local stresses caused by a restriction of the plastic zone surrounding neighboring particles. The constraints in mechanical deformation caused by the presence of the reinforcing SiC particles are largely dependent on particle size, shape and volume fraction. Neglecting particle shape as a contributing factor, all of the reinforcing particles are considered to be spherical in shape and uniformly distributed through the magnesium alloy metal matrix. The mean free path for dislocation movement between two particles is a function of particle size and volume fraction. With an increase in size of the reinforcing particle the volume fraction decreases, the Mean Free Path (say ‘q’) increases and the overall constraints on mechanical deformation of the composite microstructure is lowered. In fact, the constraints imposed on deformation are proportional to the inverse of mean free
path. In the presence of constraints in deformation the local strain to cause fracture of the reinforcing SiC particle can be expressed as (2) 3fracture ¼ 30 CKq where 3O is the matrix strain in the absence of reinforcing particles, C is the constraints imposed by the reinforcing SiC particulates [28,30], and q (mean free path)ZK1. The short interparticle distance or decrease in mean free path arising from a high volume fraction of the reinforcing particles and concomitant particulate clustering facilitates rapid linkage between neighboring voids. The coalescence of the fine microscopic voids results in pockets of shallow dimples observed on the fatigue fracture surface (Fig. 10c). At a lower cyclic stress amplitude (sZ124 MPa and N fZ364,874 cycles) the region of detectable crack formation and early growth was rough on a microscopic scale (Fig. 11a) and revealed evidence of a population of
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
219
Fig. 10. Scanning electron micrographs of the fatigue fracture surface of the Z6/SiC/20p composite cyclically deformed at stress amplitude of 151 MPa, NfZ 114,490 cycles at 150 8C, showing: (a) overall morphology, (b) microscopic cracks in transgranular fracture surface, (c) voids, shallow dimples and microcracks, (d) SiC particulate cracking.
macroscopic and microscopic voids (Fig. 11b). Progressive growth of the fine microscopic voids during continued cyclic stressing, and culminating in their coalescence results in fracture. The growth of voids can be envisioned as a cavity enlargement process, which is enhanced by minimal constraints in localized plastic deformation. The coalescence of the macroscopic voids is aided by the fine microscopic voids in the microstructure (Fig. 11d). The overload fracture surface revealed evidence of SiCp cracking (Fig. 11c) and numerous microscopic cracks (Fig. 11d). The fracture surfaces revealed features reminiscent of locally ductile and locally brittle failure mechanisms, namely, microscopic voids, dimples, SiC particulate cracking, and microscopic cracks. During cyclic deformation the large mismatch in strain carrying capability that exists between the hard, brittle and elastically deforming SiC particle and the soft, ductile and plastically deforming magnesium alloy metal-matrix results in a concentration of
stress at and near the reinforcing SiC particle making it a favorable site for failure by cracking (Fig. 11c). The void nucleation and cracking of the SiCp suggests that local plastic strains dominate in regions containing a high volume fraction of the SiC particulate reinforcement. Fracture of the reinforcing SiCp, in the plastic zone ahead of an advancing crack, generates conditions conducive for crack tip extension, which in turn is favorable for enhancing crack propagation rate. For a constant volume fraction (‘f’) of the SiCp reinforcement the tendency for it to fracture by cracking is greatest when the tensile stress exceeds the fracture strength of the particle. A gradual increase in the local stress concentration favors the initiation of microcracking at the SiC particles either individually dispersed or in clusters, while concurrently favoring decohesion or separation at their interfaces with the metal matrix. Numerous microscopic cracks were found at the reinforcing SiC particulates. Besides, the matrix between
220
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
Fig. 11. Scanning electron micrographs of the Z6/SiC/20p-F composite cyclically deformed at stress amplitude of 124 MPa, NfZ364,874 cycles, showing: (a) overall morphology, (b) high magnification of (a) showing population of microscopic voids, (c) cracked SiC particulate, (d) region of overload showing microscopic cracks and dimples.
particulate clusters is subjected to a high triaxial stress, which favors fracture of the matrix while concurrently exacerbating microscopic crack initiation at the SiC reinforcing particulates. The microscopic cracks grow through the SiC particulate clusters and link by fast fracture through the magnesium alloy metal matrix to adjacent SiC particles. This results in a macroscopic crack. Rapid propagation of the macroscopic cracks along with the microscopic cracks is responsible for the inferior cyclic stress resistance at ambient temperature (27 8C). Particulate fracture aided by decohesion at its interfaces with the metal matrix, void nucleation, growth and coalescence serves to increase the rate of microcrack growth by providing conditions conducive for crack-tip extension. For a constant volume fraction of the SiC particulate reinforcement, the probability for the reinforcing particle to fracture increases at the higher cyclic stress amplitudes and concomitant shorter fatigue life.
5. Conclusions A comprehensive study of the tensile deformation, cyclic stress-controlled fatigue and final fracture (quasi-staticC cyclic) behavior of Z6/SiCp metal-matrix composite provides the following highlights: 1. The initial microstructure of the Z6/SiC/20p-F metal matrix composite revealed a near-uniform distribution of the reinforcing SiC particulates in the longitudinal direction of the extruded composite plate. At intervals, an agglomeration or clustering of the reinforcing SiC particulates was observed. 2. The presence of hard, brittle and elastically deforming SiC particulates in a soft, ductile and plastically deforming magnesium alloy metal matrix favors the initiation of fine microscopic cracks at low values of applied stress. Extensive fractography of the failed
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
samples revealed limited ductility on a macroscopic scale, but microscopic features were reminiscent of locally ductile and brittle mechanisms. Fracture of the metal matrix between the dispersed SiC particulates and particulate clusters coupled with failure of the reinforcing SiC particulate by cracking facilitates rapid growth of the microscopic cracks and eventual coalesce by fracture through the ductile metal matrix. This results in macroscopic failure and low tensile ductility. 3. Influence of temperature on stress amplitude–fatigue life response revealed degradation in cyclic fatigue life with an increase in test temperature from 27 to 150 8C. When the cyclic stress amplitude is normalized with respect to yield strength (sYS) and ultimate tensile strength (sUTS) of the material at the test temperature, the variation with fatigue life reveals an improved cyclic stress resistance at 150 8C. The observed improvement in cyclic fatigue life is more pronounced at the higher ratios of stress than at the lower ratios of stress. 4. At a given temperature, cyclic fracture morphology was found to be essentially similar over the range of cyclic stress amplitudes. Macroscopic observations revealed the fracture surface to be brittle, i.e.: low cyclic ductility, at all cyclic stress amplitudes, but having microscopic features reminiscent of locally ductile and brittle mechanisms. Constraints in mechanical deformation, induced in the metal matrix, by the hard, brittle and elastically deforming SiCp reinforcement phase coupled with local stress concentration effects at the matrix-reinforcement particle interfaces promote failure through the conjoint influences of particle cracking and fast fracture through the matrix. Fracture of the plastically deforming matrix occurred through the formation, growth and coalescence of the voids.
Acknowledgements This research was jointly supported by State of Ohio: Board of Regents (Columbus, OH, USA) (Grant No. R4414OBR), the University of Akron (Akron, OH, USA) and personal funds of Dr Srivatsan. Sincere thanks and appreciation are extended to Dow Chemical Company (Freeport, TX, USA) for providing the material used in this study. Sincere thanks and abundance of appreciation is extended to the unknown reviewers whose comments, suggestions and corrections to text have most certainly helped strengthen the overall manuscript. One of the authors (Dr TSS) extends profound thanks to Mr G. Guruprasad (Graduate Student: Department of Mechanical Engineering) for terminal assistance with recompiling the manuscript for re-submission.
221
References [1] Lewandowski JJ. In: Clyne TW, editor. Metal matrix composites, vol. 3. Amsterdam: Elsevier; 2000. p. 151–87. [2] Divecha AP, Fishman SG, Karmarkar SD. Synthesis of metal-matrix composites. J Metals 1981;33:12–15. [3] Stephens JR. High temperature metal matrix composites for future aerospace systems. NASA TM 1987;100–212. [4] Nair SV, Tien JK, Bates RC. Int Metals Rev 1985;30(6):285–97. [5] Srivatsan TS, Sudarshan TS, editors. Rapid solidification technology: an engineering guide. Lancaster, PA: Technomic; 1993. p. 603–700. [6] Koczak MJ, Khatri SC, Allison JE, Bader MG. In: Suresh S, Mortensen A, Needleman A, editors. Fundamentals of metal matrix composites. Boston, MA: Butterworth/Heinemann; 1993. p. 297–317. [7] Taya M, Arsenault RJ. Metal matrix composites: thermomechanical behavior. Elmsford, NY: Pergamon Press; 1989. [8] McDanels DL. Metall Trans 1985;16A:1105–15. [9] Allison JE, Jones JW. Fatigue behavior of discontinuously-reinforced metal-matrix composites. In: Suresh S, Mortensen A, Needleman A, editors. Fundamentals of metal matrix composites. London: Heinemann; 1993. p. 269–96. [10] Srivatsan TS, Sudarshan TS, Lavernia EJ. Processing of discontinuously reinforced composite materials by rapid solidification. Prog Mater Sci: Int Rev J 1995;39(4/5):317–409. [11] Christman T, Suresh S. Microstructural development in an aluminum alloy-SiC whisker composite. Acta Metall 1988;36(7): 1699–704. [12] Llorca J, Needleman A, Suresh S. An analysis of the effects of matrix void growth on deformation and ductility in metal-ceramic composites. Acta Metall 1991;39(10):2317–35. [13] Christman T, Suresh S. Mater Sci Eng 1988;102:211–20. [14] Christman T, Needleman A, Suresh S. An experimental and numerical study of deformation in metal-ceramic composites. Acta Metall 1988; 37:211–20. [15] Hunt Jr. WH. Discontinuously-reinforced aluminum: the second generation. In: Ravi VA, Srivatsan TS, Moore JJ, editors. Processing and fabrication of advanced materials III. Warrendale, PA: TMS: The Minerals, Metals and Materials Society; 1994. p. 663–90. [16] Mikucki BA, Mercer II WE, Green WG. Extruded magnesium alloys reinforced with ceramic particles. Annual Meeting of ASM International; 1990. [17] Chaudhury PK, Rack HJ, Mikucki BA. J Mater Sci 1991;26: 2343–7. [18] Couling SL. Use of air/CO/SF mixtures for improved protection of molten magnesium. Paper presented at the 36th Annual World Conference on Magnesium, Oslo, Norway; 24–28 June 1979. p. 54. [19] ASTM Standard E-8. Standard Test Method for Tension Testing of Metallic Materials. American Society for Testing and Materials, Race Street, Philadelphia, PA, USA; 1993. [20] Shang JK, Ritchie RO. Crack bridging by uncracked ligaments during fatigue crack growth in silicon carbide reinforced aluminum alloy composite. Metall Trans 1989;20A:897–908. [21] Osman TM, Lewandowski JJ, Hunt Jr WH. Microstructure-property relationships for an aluminum/SiCp composite with different deformation histories. In Processing and fabrication of fiber-reinforced MMCs. Proceedings of an international conference, Montreal. Materials Park, OH, USA: American Society for Materials International; September 1990. p. 209–16. [22] Srivatsan TS, Mattingly J. Influence of heat treatment on the tensile behavior of an aluminum alloy-ceramic particle composite. J Mater Sci 1993;28:611–20. [23] Srivatsan TS. Influence of secondary processing on cyclic fracture behavior of an aluminum-ceramic particle composite. In: Ravi VA,
222
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
Srivatsan TS, Moore JJ, editors. Processing and fabrication of advanced materials III. TMSD, Warrendale, PA: TMS: The Minerals, Metals and Materials Society; 1994. p. 699–731. [24] McClintock FA. Ductility. Materials Park, OH, USA: American Society for Metals; 1968 p. 256–66. [25] Van Stone RH, Cox TB, Low Jr JR, Psioda JA. The fracture behavior of metals. Int Metals Rev 1975;30:157–71. [26] Argon AS, Im J, Safoglu R. Cavity formation from inclusions in ductile fracture. Metall Trans 1975;6A:825–31.
[27] Lloyd DJ. Aspects of fracture in particulate-reinforced metal-matrix composites. Acta Metall Mater 1991;39(1):59–71. [28] Derby B, Mummery PM. In: Suresh S, Mortensen A, Needleman A, editors. Fundamentals of metal-matrix composites. Boston, MA: Butterworth/Heinemann; 1992. p. 251–71. [29] Lloyd DJ. Int Mater Rev 1994;39:1–23. [30] Haasen DF, Crowe CR, Ahearn JS, Cooke DC. In: Early JG, Shives TR, Smith JH, editors. Failure mechanisms in high performance materials, 1984. p. 147.
The quasi-static, cyclic fatigue and final fracture behavior of a magnesium alloy metal-matrix composite T.S. Srivatsan*, Meslet Al-Hajri, P.C. Lam Division of Materials Science and Engineering, Department of Mechanical Engineering, College of Engineering, University of Akron, Akron, OH 44325-3903, USA Received 10 April 2004; revised 23 September 2004; accepted 23 September 2004 Available online 23 November 2004
Abstract In this paper the results of a study to understand the role of composite microstructure on quasi-static, cyclic fatigue, deformation and fracture behavior of magnesium alloy Z6 discontinuously reinforced with silicon carbide (SiC) particulates are presented. Quasi-static fracture of the composite was dominated by cracking of the reinforcing particulates present in the magnesium alloy metal matrix. Final fracture occurred as a result of crack propagation through the alloy matrix between particulate clusters. The composite specimens were cyclically deformed over a range of stress amplitudes. The stress amplitude–fatigue life response was found to degrade with an increase in test temperature. Influence of test temperature on quasi-static and cyclic response and final fracture behavior of the composite is discussed in light of the concurrent and mutually interactive influences of composite microstructural effects, deformation characteristics of constituents of the composite material, magnitude of cyclic stress amplitude and fatigue life. q 2004 Elsevier Ltd. All rights reserved. Keywords: Magnesium alloy; Particulate reinforcement; Composite; Microstructure; Fatigue; Fracture; Temperature
1. Introduction A growing and sustaining interest arising primarily from the need to improve fuel economy, minimize vehicle emissions, enhance styling options, improve overall performance, while concurrently maintaining safety, quality, reliability, durability while also ensuring profitability, are some of the challenges that most certainly need to be addressed by emerging and future generation materials that find use in a variety of performance-critical applications in the industries spanning aerospace, automotive and ground transportation. This burgeoning need has provided the impetus for the development, emergence and rapidly increasing use of a ‘new’ class of metallic materials, namely, metal-matrix composites (MMCs) [1–4]. Their exceptionally high specific modulus (E/r), strength-toweight (s/r) ratios, fatigue strength, wear resistance, and tailorable properties such as thermal expansion coefficient,
* Corresponding author. Tel.: C1 330 972 6196; fax: C1 330 972 6027. E-mail address: [email protected] (T.S. Srivatsan). 1359-8368/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.compositesb.2004.09.004
are some of the important attributes that have helped make MMCs attractive and potentially viable candidates for selection and use in several weight-sensitive, stiffnesscritical and performance-worthy components in both air and ground transportation systems [2–8]. Driven initially by the rising demand for high performance military and space applications, the discontinuously reinforced metal matrices have most certainly stimulated considerable scientific and technological interest in the last two decades. Rapid strides in recent years have seen their use for a wide range of applications, specifically recreational [2,4,8], automotive [9] and some aerospace products [3,4]. Also, the emerging need for satisfying high fuel-economy goals in the automobile industry will be a major challenge for which the metal-matrix composites based on the lightweight alloys will play a significant role. However, problems arising from inadequate fracture toughness, limited damage tolerance, poor tensile ductility, inferior fracture-related properties, size limitations on available product forms, and intrinsic material variability, continue to pose a severe limitation to the wide spread application and use of the metal-based composites [2,4,5,10].
210
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
A class of metal-matrix composite materials are attracting use in several automotive components and consumerrelated applications, is the magnesium alloy matrices discontinuously reinforced with particulate, whisker or short-fiber reinforcements. These discontinuously reinforced metal matrices will be referred to henceforth in this paper as DRMg MMCs. The presence of a discontinuous reinforcement phase in a continuous magnesium alloy metal matrix can result in properties not attainable by other means, thus extending the possible applications [7]. Both die cast and extruded Mg/SiC particle composites exhibit attractive physical properties. In fact die cast bars of AZ91/ SiC have shown an increase in elastic modulus, offset yield strength and ultimate tensile strength as the percentage of SiC reinforcement increases [10]. At room temperature the metal matrix reinforced with 25% SiC by volume had an elastic modulus 72% higher than that of the unreinforced alloy. The yield strength was 47% greater and the ultimate tensile strength was 23% greater than the unreinforced counterpart. With an increase in temperature, the properties deteriorated, but in the same manner for both the reinforced metal matrix and the unreinforced counterpart. An ability to achieve improvements in mechanical properties is largely dependent on the mutually interactive influences of: (a) intrinsic properties of the composite constituents, i.e. metal matrix and reinforcement, and (b) the size, shape, orientation, volume fraction and distribution of the reinforcement phase in the metal matrix [11–14]. Selection of type of reinforcement and its geometry (shape) is critical to obtain the optimum combination of properties at a substantially low cost [15]. Of the available choices silicon carbide particulates (SiCp) are the preferred reinforcements primarily because enhanced properties can be achieved with little or no penalty in density [9]. From an economic standpoint, the use of ceramic particulates as the reinforcement phase is advantageous. The SiC particulate reinforcement is easily available and relatively inexpensive. The magnesium alloy metal matrix reinforced with ceramic particles is a promising material based on its potential for low cost [16]. From a design engineering perspective, the attractiveness in preferring and choosing a discontinuously reinforced magnesium alloy MMC for many applications stems from an improvement in specific modulus, that is, the density compensated increase in elastic modulus. The discontinuously reinforced magnesium alloy metal matrices (DRMg MMCs) can be tailored so as to have properties that are near isotropic. Furthermore, discontinuously reinforced magnesium alloy metal matrices maintain their amenability to: (a) conventional metallurgical processing, (b) fabrication, and (c) characterization methods currently used for the unreinforced counterparts [2,4,5,10]. Several practical applications for the DRMg MMCs involve cyclic loading, and therefore knowing the fatigue response coupled with an understanding of the resultant fracture characteristics is essential. Reinforcing a magnesium alloy metal matrix can have a beneficial or
detrimental influence on cyclic fatigue resistance depending upon diverse and yet related factors such as: (i) method of synthesis (primary processing method), (ii) reinforcement type, (iii) geometrical constitution (size, shape, volume fraction, and distribution of the reinforcing phase), (iv) nature of secondary processing (aging treatments), and (v) strength of the metal matrix and reinforcement/matrix interfaces. The presence of discontinuous particulate reinforcements in a magnesium alloy metal matrix does exert an influence on crack growth behavior, i.e. extent and severity of damage propagation, and resultant fracture characteristics. The influence on crack growth, or damage propagation, is governed by the mutually interactive influences of matrix composition and microstructure, geometrical constitution, processing conditions and applied cyclic stress intensity range. While few studies have focussed on developing an understanding of the nature and influence of the reinforcement particles on matrix microstructure and tensile response, the fatigue database for discontinuously reinforced magnesium alloy matrices is limited, and there certainly exists a need to expand this aspect of mechanical characterization. Presence of discontinuous particulate reinforcements in a ductile magnesium alloy metal matrix will alter precipitation kinetics of the material during heat treatment (i.e. aging treatment) and thus the microstructure of the composite compared with the unreinforced monolithic counterpart [13,17]. A change in intrinsic microstructural features will exert an appreciable influence on mechanical response and resultant fracture behavior. The objective of this paper is to understand the influence of discontinuous particulate reinforcement on tensile, cyclic fatigue and fracture (quasi-static and cyclic) behavior of a magnesium–zinc alloy reinforced with fine particulates of silicon carbide (henceforth referred to as SiCp). The tensile deformation, fatigue response and fracture behavior of the composite are discussed in light of the concurrent and mutually interactive influences of composite microstructural effects, matrix deformation characteristics, test temperature and nature of loading.
2. Material Elemental magnesium was melted in a stainless steel crucible. The melt was prevented from burning by using a protective atmosphere of carbon dioxide, air and SF6 [18]. A nominal 6.0 wt% of zinc was added to the melt. This was followed by the addition of silicon carbide (SiC) particulates (mesh size: 600 grit) to the molten magnesium–zinc alloy using proprietary techniques. The molten composite material was cast into billets. During casting a stainless steel impeller was used to continuously stir the composite melt so as to retain the SiC particulates in suspension. Prior to extrusion, the composite billet was homogenized at 672 K for 4 h followed by quenching in cold water. The billet was
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
extruded into a solid rod using extrusion techniques similar to that used for conventional magnesium. The reduction ratio used was 60:1. The manufacturer (DOW Chemical Company at Freeport, TX, USA) considers precise details of the processing treatments (primary and secondary) as proprietary. The Z6/SiC/20p composite material was provided in the as-extruded condition (denoted henceforth by the symbol ‘F’).
3. Experimental techniques 3.1. Specimen preparation Blanks of length 150 mm were cut from the as-received Z6/SiC/20p-F composite extrusion using a diamond-coated blade. Smooth cylindrical test specimens (gage lengthZ 6.25 mm and gage diameterZ25 mm) were precision machined from the blanks using diamond tooling. The specimens were machined with the stress axis parallel to the extrusion direction and conformed to specifications outlined in ASTM Standards E8-93 [19] and E606-93 [20]. The length-to-diameter ratio of the mechanical test specimen was chosen so as to minimize buckling during fully reversed (tension–compression) stress amplitude-controlled cyclic deformation. To minimize the effects of surface irregularities and finish, the test specimen surface was prepared by mechanically polishing the gage section using progressively finer grades of silicon carbide impregnated emery paper and then finish polished using 0.5 mm alumina powder suspended in distilled water so as to obtain a mirror like finish that is free of all circumferential scratches and surface machining marks. 3.2. Initial microstructure characterization Metallographic samples were cut from the as-received composite extrusion, mounted in bakelite and wet ground on 320, 400 and 600 grit SiC impregnated emery paper, using copious amounts of water as lubricant. The mounted and ground samples were then mechanically polished using a 1 mm alumina-powder suspended in distilled water. Fine polishing to near mirror like finish was achieved using 0.5 mm diamond paste. Reinforcement morphology and its distribution in the metal matrix along with other intrinsic microstructural features were identified by examining the samples in an optical microscope and photographed using bright-field illumination technique.
211
The tests were conducted in controlled laboratory air environment (relative humidity of 55%) at ambient temperature (27 8C). Tensile tests were conducted in accordance with procedures outlined in ASTM E8 at a strain rate of 1!10K4 sK1 and using a 12.5 mm gage length extensometer to measure strain. Reduction in area was measured at the fracture surface of the failed specimen using a micrometer. In considering the limited tensile ductility, cyclic stress resistance, and high-cycle fatigue (HCF) life of the silicon carbide particulate reinforced magnesium alloy at room (ambient) temperature, the stress amplitude-controlled highcycle fatigue tests were also conducted at an elevated temperature. The elevated temperature chosen was 150 8C and corresponded to the maximum temperature limit of the environmental chamber unit. The elevated temperature tests were conducted inside an environmental chamber unit (Instron: Model 3111). The temperature was controlled with the aid of a temperature controller in conjunction with a thermocouple. The thermocouple was fixed onto the specimen’s surface. Maximum temperature variation was well within 2 8C of the set-point temperature (150 8C) over the entire duration of the test. Before each test, the specimen was maintained at the test temperature for 30 min so as to achieve stability with the environment. The fatigue tests were conducted at a constant cyclic frequency of 5 Hz and at a stress ratio (RZsminimum/smaximum) of 0.1. The number of cycles to cause complete failure or separation is taken as fatigue life (Nf). 3.4. Failure-damage analysis Fracture surfaces of the tensile and cyclically deformed test samples were examined in a scanning electron microscope (SEM) to: (a) determine the macroscopic fracture mode, and (b) characterize the fine-scale topography to establish the microscopic mechanisms governing fracture. The distinction between macroscopic mode and microscopic fracture mechanisms is based on the magnification level at which the observations are made. The macroscopic mode refers to the nature of failure, while the microscopic mechanism relates to the local failure processes (microscopic void formation, coalescence, and nature and intensity of cracking). Samples for SEM observation were obtained from the failed specimens by sectioning parallel to the fracture surface. Matching fracture surfaces were viewed to determine the presence of fractured SiCp on both halves of the specimen.
3.3. Mechanical testing 4. Results and discussion Test specimens were precision machined in accordance with procedures outlined in standards ASTM E8 and E606. The tensile and cyclic fatigue tests were performed on a fully automated, closed-loop servohydraulic test machine (INSTRON) equipped with a 10,000 kg (98 kN) load cell.
4.1. Undeformed microstructure Fig. 1 is the optical micrograph showing microstructure of the Z6/SiC/20p composite The silicon carbide
212
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
Fig. 1. Optical micrograph illustrating microstructure of the Z6/SiC/20p-F composite.
particulates (SiCp), in the Z6 magnesium alloy metal matrix, were generally of near-uniform size. However, a few irregularly shaped particulates were found dispersed randomly between the near uniform shaped particulates. An agglomeration or clustering of the SiCp was observed. An agglomerated site consisted of the larger SiC particulates intermingled with the smaller, uniform and more regularly shaped SiC particles. Alignment of the reinforcing SiC particulates was observed in the direction of billet extrusion. Such an alignment has been documented in other extruded aluminum alloy-based MMCs [21–24]. However, in this study no attempt was made to determine the particle size distribution for the Z6/SiC/20p-F MMC material. The etched composite microstructure revealed the matrix to consist of very fine grains, which could not be easily resolved in the optical microscope at magnifications !1000!. 4.2. Tensile properties and fracture The tensile properties of the Z6/SiC/20p-F composite, at the two test temperatures, are summarized in Table 1. The results are the mean values based on duplicate tests. At ambient temperature (27 8C), the composite material revealed limited ductility, quantified in terms of both
elongation-to-failure (3f) and reduction-in-area (RA). An increase in test temperature resulted in a 46% decrease of yield strength (sys) and a 50% drop in ultimate tensile strength (sUTS) but a concomitant improvement in: (a) elongation-to-failure (3f) by over 800%, and (b) the reduction in area by 500%. It is evident that the degradation in strength had a profound influence on tensile ductility of this composite. The tensile fracture surfaces provide valuable information pertaining to microstructural effects on tensile ductility and fracture properties of the magnesium metal matrix composite. It is fairly well documented that the fracture of unreinforced monolithic alloys occurs by events of microscopic void nucleation and growth, with the nucleation essentially occurring at the coarse intermetallic particles and other insoluble phases in the microstructure [25,26]. An essential requirement for void nucleation either at a reinforcing particle and/or other coarse second-phase particles in the microstructure is the development of a critical normal stress across the particulate-matrix interfaces [27]. In the unreinforced metal matrix, the nucleation of cavities and voids is favored by the concurrent and mutually interactive influences of: (1) cracking of the hard, intrinsically brittle and elastically deforming SiCp inclusions, and (2) decohesion at interfaces between the hard and brittle reinforcing SiC particulate and the soft and ductile magnesium alloy metal matrix. The Z6/SiC/20p-F composite exhibited limited ductility at ambient temperature, with fracture, on a macroscopic scale, occurring on a plane normal to the far-field stress axis. High magnification observations of the tensile fracture surface revealed microscopic features reminiscent of locally ductile and brittle mechanisms. Representative fractographs of the tensile fracture surface at the two temperatures are shown in Figs. 2–4 and discussed in the following sections. 4.2.1. Test temperature of 27 8C Tensile fracture of the Z6/SiC/20p-F MMC at room temperature was flat, near featureless and essentially on a plane normal to the far field stress axis when viewed on a macroscopic scale (Fig. 2a). Microscopic observations revealed cracking parallel to the major stress axis (Fig. 2b). The macroscopic cracks were surrounded by cracked particulates and ductile regions, which are referred to as tear ridges. Isolated tear ridges were evident on the tensile fracture surface (Fig. 2c). The matrix was covered with numerous microscopic cracks (Fig. 2d) and cracked SiC particulates (Fig. 3a). The overload region revealed
Table 1 Tensile properties of Mg-6Zn/SiC/20p-F composite Temperature (8C)
Elastic modulus (GPa)
27 150
66 48
Yield strength
Ultimate tensile strength
MPa
ksi
MPa
ksi
307 166
44.5 24
379 186
55 27
YS/UTS (%)
Elongation (%)
Reduction in area (%)
ln(A0/Af) (%)
81 89
2.0 19.0
3.0 19.0
3.2 21.0
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
213
Fig. 2. Scanning electron micrographs of the tensile fracture surface of Z6/SiC/20p-F composite at ambient temperature (27 8C).
Fig. 3. Scanning electron micrograph of the tensile fracture surface of the Z6/SiCp composite showing: (a) cracked SiCp, and (b) fine microvoids and isolated shallow dimples.
214
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
a population of microscopic voids of varying size and shapes intermingled with ductile dimples (Fig. 3b), features reminiscent of locally ductile fracture. 4.2.2. Test temperature of 150 8C On a macroscopic scale, the tensile fracture surface of the Z6/SiC/20p-F MMC deformed at the elevated temperature (150 8C) was flat and similar to the ambient temperature fracture surface but relatively rough when viewed on a microscopic scale (Fig. 4a). Isolated pockets of macroscopic cracks were distributed randomly through the fracture surface (Fig. 4b). The fracture surfaces revealed the presence of cracked SiC particulates surrounded by a population of shallow dimples. The matrix of the composite was covered with a population of microscopic voids of a wide range of sizes (Fig. 4c). The voids were distributed throughout the fracture surface and were intermingled with shallow dimples (Fig. 4d). The constraints in mechanical deformation imposed by the hard, intrinsically brittle and elastically deforming SiCp
in a relatively soft and ductile magnesium alloy metal matrix coupled with the concurrent development of a triaxial stress state in the matrix of the Z6/SiCp composite is a factor that strongly aids in limiting flow stress of the composite microstructure, while concurrently placing constraints on microscopic void nucleation and growth. As a direct consequence of the deformation constraints induced by the discontinuous SiC particulates, a higher applied stress is required to initiate plastic deformation in the microstructure of the metal-based composite. This translates to a higher elastic stiffness and yield strength of the Z6/SiC/20p-F composite when compared to the unreinforced Z6 alloy counterpart. Failure of the reinforcing SiC particulates either by cracking or decohesion at their interfaces with the magnesium alloy metal matrix can be attributed to the mutually interactive influences of the following: (a) local plastic constraints, (b) particle size, and (c) degree of particle agglomeration. The local plastic constraints are particularly important for the larger-sized SiC particles and particle
Fig. 4. Scanning electron micrographs of the tensile fracture surface of Z6/SiC/20p-F composite at 150 8C.
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
clusters during composite fracture [28–30]. Examination of the tensile fracture surfaces revealed the damage associated with fracture to be localized at the discontinuous SiCp reinforcements with little evidence of void formation away from the fractured SiC particulates. Fracture of the hard, brittle and elastically deforming SiCp was found to be greater in regions of particle clustering. This is attributed to: (1) enhanced local stresses resulting from restriction of plastic deformation, and (2) an intrinsic brittleness of the reinforcing SiCp with the propensity for it to fracture due to localized deformation. The two concurrent and mutually competitive factors result in: (a) particulate failure by cracking, and (b) aided by decohesion or separation at its interfaces with the metal matrix, being the dominant damage modes. The damage of the composite microstructure, during uniaxial loading, arising from the synergistic influences of particle cracking and decohesion at the magnesium alloy matrix–SiC particulate interfaces results in a detrimental influence on tensile ductility. Furthermore, assuming that the metal matrix-reinforcement particle interfaces are strong, the triaxial stresses generated during far-field tensile loading favor limited growth of the microscopic voids in the matrix of the composite. Under the influence of an applied tensile load, the microscopic voids appeared to have undergone limited growth confirming a possible contribution from SiC particle constraint-induced triaxiality on deformation and failure of the composite matrix. The limited growth of the microscopic voids without their coalescence, as a dominant fracture mode for the Z6/SiC/20p-F6 composite, suggests that the deformation properties of the magnesium alloy metal matrix are significantly altered by the presence of discontinuous SiCp reinforcements. Fracture of the brittle SiC particulates aided by failure, or decohesion at the matrix–particulate interfaces results in the formation of voids. Few of the fine microscopic voids coalesce and the halves of these voids are the shallow dimples observed on the tensile fracture surface (Fig. 3b). The lack of formation of dimples, as a dominant fracture mode, is attributed primarily to the constraints in plastic flow of the composite matrix caused by the presence of the discontinuous SiCp reinforcement phase and not due to limited ductility of the Z6 magnesium alloy matrix. The constraint on plastic flow favors the formation of fine tear ridges between the SiC particulates in the magnesium alloy metal matrix.
215
Fig. 5. Influence of temperature on cyclic stress amplitude (Ds/2)-fatigue life (Nf) response of the Z6/SiC/20p-F composite.
essentially unaffected during stress controlled cyclic deformation. The results of the axial stress amplitude controlled tests are shown in Fig. 5 as the variation of stressamplitude (Ds/2) with fatigue life (cycles to failure NF). The best curve was drawn through the test data points. Coincidentally, the stress amplitude (Ds/2)–fatigue life (Nf) curve exhibits the general trend shown by the non-ferrous metals, i.e. increasing fatigue life with decreasing stress amplitude levels. At equivalent stress amplitudes an increase in test temperature resulted in degradation in cyclic fatigue life by at least an order of magnitude. To understand and establish the reasons responsible for the observed degradation in life, the stress amplitude–fatigue life data are replotted in terms of elastic strain amplitude in the manner suggested by Hassen and co-workers [30], where at a given test temperature the elastic strain amplitude (D3e/2) is the stress amplitude (Ds/2) at that temperature normalized by elastic modulus (E) at that temperature. Variation of elastic strain amplitude with fatigue life is shown in Fig. 6. This accounts for the decrease in stiffness
4.3. High cycle fatigue behavior The stress amplitude-controlled high-cycle fatigue test provides information pertinent to the existence of an endurance limit. In the low stress amplitude and resultant high cyclic fatigue life of Mg MMC the property changes are expected to occur at and around the vicinity of the fracture process zone, i.e. the small region immediately ahead of an advancing crack tip where crack extension originates, while the material in the far field remains
Fig. 6. Elastic strain amplitude (D3e/2)–fatigue life (Nf) response of the Z6/SiC/20p-F composite.
216
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
of the Z6/SiC/20p-F composite material with an increase in test temperature. A minimal difference in elastic strain amplitude versus fatigue life behavior is observed at the two test temperatures suggesting that the demonstrated degradation in cyclic fatigue life, with an increase in test temperature, is a reflection of the lower elastic modulus (48 GPa) of the composite material at the higher test temperature (150 8C) when compared with the modulus (66 GPa) at ambient temperature (27 8C). To further examine the influence of temperature on stress amplitude controlled fatigue response, the stress–fatigue life data are replotted to account for differences in strength of the composite material at the two temperatures. For purposes of rationalizing influence of composite material strength (yield strength versus ultimate tensile strength) on stress amplitude versus fatigue life response the stress amplitude was normalized with strength at that temperature so as to evaluate the feasibility of using the composite in performance-critical applications. In Fig. 7A the stress amplitude (Ds/2) has been normalized by the uniaxial yield strength
Fig. 7. (A) Variation of cyclic stress amplitude normalized with yield strength (sYS) versus fatigue life (Nf) for the Z6/SiC/20p-F composite at the two temperatures. (B) Variation of cyclic stress amplitude normalized with ultimate tensile strength (sUTS) versus fatigue life (Nf) for the Z6/SiC/20p-F composite at the two temperatures.
(sYS) of the composite at the temperature, while in Fig. 7B the cyclic stress amplitude (Ds/2) has been normalized by the ultimate tensile strength. (sUTS) of the composite at the temperature. In both cases the normalized stresses are plotted against the number of cycles-to-failure (Nf). In both situations the Z6/SiC/20p-F composite material reveals improved high cycle fatigue life at the higher test temperature. The improvement is significant at the higher ratios of stresses and resultant low fatigue life than at the lower ratios of stresses and resultant enhanced fatigue life. The beneficial effects of elevated temperature on cyclic fatigue resistance at the higher ratios of stress are a direct indication of the superior crack initiation and/or crack growth resistance of the composite material at the elevated temperature. 4.4. Cyclic fracture behavior Examination of the fracture surfaces of the cyclically deformed fatigue specimens, in a JEOL scanning electron microscope, was done at: (a) low magnification to identify the regions of fatigue initiation and final fracture (overload), and (b) at higher magnifications in the fatigue region to identify the regions of microcrack formation and early microscopic crack growth, and the overload region to identify the fine-scale fracture features. Representative fracture features of the Z6/SiC/20p-F composite are shown in Figs. 8–11. 4.4.1. Ambient temperature (27 8C) High cycle fatigue fracture of the composite sample deformed at the maximum cyclic stress of 283 MPa (NfZ6369 cycles) revealed fracture to have essentially occurred on a plane that is normal to the far-field stress axis. The fracture surface revealed a small region of fatigue, about the size of a thumbnail, and a large portion to be overload failure (Fig. 8a) with little evidence of gross cyclic ductility. Low magnification observations revealed randomly distributed microscopic cracks in the region of early crack growth (Fig. 8b); features reminiscent of locally brittle failure. The microscopic cracks were common in test specimens deformed at the higher cyclic stress and resultant short fatigue life. There was little evidence of crack branching suggesting that the damaged region ahead of the crack tip be greatly reduced. High magnification observations of the overload region revealed a population of fine microscopic voids of varying size and shapes and shallow dimples (Fig. 8c) and cracked SiC particulates (Fig. 8d), features reminiscent of locally ductile and brittle failure. The fine microscopic voids were distributed through the fracture region with the occurrence of microvoid coalescence. The composite sample cyclically deformed at the lower stress of 184 MPa with resultant enhanced fatigue life (NfZ 162,197 cycles) revealed distinct regions of stable crack growth and overload (Fig. 9a). High magnification
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
217
Fig. 8. Scanning electron micrographs showing HCF fracture surface features of the Z6/SiC/20p-F composite deformed at cyclic stress amplitude of 283 MPa, NfZ6369 cycles, showing: (a) overall morphology, (b) array of microscopic cracks, (c) voids of varying size and shallow dimples on the overload fracture surface, (d) cracked of SiC particulate.
observations in the region of stable crack growth revealed it to be flat with an array of fine microscopic cracks (Fig. 9b). The overload region revealed microscopic cracks, isolated pockets of shallow dimples and microscopic voids (Fig. 9c), features reminiscent of locally brittle and ductile failure. The randomly distributed voids revealed evidence of microvoid coalescence (Fig. 9d). The microscopic voids ranged in size from the sub micron to tens of microns. 4.4.2. Elevated temperature (150 8C) At this temperature the fatigue fracture surface of the composite sample cyclically deformed at the lower stress amplitude and resultant longer fatigue life (Fig. 11) was quite similar to that observed at the higher stress amplitude and short fatigue life (Fig. 10). In addition to cracking of the reinforcing particulates, decohesion or separation, promoted by local stress concentration arising from dislocation buildup during cyclic deformation, was evident at the metal
matrix–particulate interfaces. The larger SiC particles fracture at a much lower stress than the smaller SiC particles. This is because the intrinsic flaw size in a reinforcing SiC particle is directly proportion to its size. Assuming flaw size (say ‘d’) is a fraction (say ‘x’) of the diameter of the particle (2r), the stress essential to fracture or rupture the second-phase/reinforcing particle is sfracture Z Kpfracture =ð2pDÞ0:5 p
(1)
where Kpfracture is the fracture toughness of the reinforcing SiC particle. Further, the larger dimensions of the reinforcing SiCp facilitate a greater degree of stress transfer from the plastically deforming metal matrix to the elastically deforming particulate. The smaller SiCp do not crack easily. However, on account of intrinsic strain differences between the plastically deforming matrix and the elastically deforming particle, the process culminates in separation, or decohesion, at its interfaces with the metal
218
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
Fig. 9. Scanning electron micrographs showing HCF fracture surface features of the Z6/SiC/20p-F composite, deformed at cyclic stress amplitude of 184 MPa, NfZ162,197 cycles, showing: (a) overall morphology, (b) high magnification of (a) showing randomly distributed microscopic cracks on transgranular fracture surface, (c) microcrack, dimples and voids, (d) void coalescence.
matrix. The tendency for SiC particulate fracture is higher in regions of high local particulate volume fraction. This arises from increased local stresses caused by a restriction of the plastic zone surrounding neighboring particles. The constraints in mechanical deformation caused by the presence of the reinforcing SiC particles are largely dependent on particle size, shape and volume fraction. Neglecting particle shape as a contributing factor, all of the reinforcing particles are considered to be spherical in shape and uniformly distributed through the magnesium alloy metal matrix. The mean free path for dislocation movement between two particles is a function of particle size and volume fraction. With an increase in size of the reinforcing particle the volume fraction decreases, the Mean Free Path (say ‘q’) increases and the overall constraints on mechanical deformation of the composite microstructure is lowered. In fact, the constraints imposed on deformation are proportional to the inverse of mean free
path. In the presence of constraints in deformation the local strain to cause fracture of the reinforcing SiC particle can be expressed as (2) 3fracture ¼ 30 CKq where 3O is the matrix strain in the absence of reinforcing particles, C is the constraints imposed by the reinforcing SiC particulates [28,30], and q (mean free path)ZK1. The short interparticle distance or decrease in mean free path arising from a high volume fraction of the reinforcing particles and concomitant particulate clustering facilitates rapid linkage between neighboring voids. The coalescence of the fine microscopic voids results in pockets of shallow dimples observed on the fatigue fracture surface (Fig. 10c). At a lower cyclic stress amplitude (sZ124 MPa and N fZ364,874 cycles) the region of detectable crack formation and early growth was rough on a microscopic scale (Fig. 11a) and revealed evidence of a population of
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
219
Fig. 10. Scanning electron micrographs of the fatigue fracture surface of the Z6/SiC/20p composite cyclically deformed at stress amplitude of 151 MPa, NfZ 114,490 cycles at 150 8C, showing: (a) overall morphology, (b) microscopic cracks in transgranular fracture surface, (c) voids, shallow dimples and microcracks, (d) SiC particulate cracking.
macroscopic and microscopic voids (Fig. 11b). Progressive growth of the fine microscopic voids during continued cyclic stressing, and culminating in their coalescence results in fracture. The growth of voids can be envisioned as a cavity enlargement process, which is enhanced by minimal constraints in localized plastic deformation. The coalescence of the macroscopic voids is aided by the fine microscopic voids in the microstructure (Fig. 11d). The overload fracture surface revealed evidence of SiCp cracking (Fig. 11c) and numerous microscopic cracks (Fig. 11d). The fracture surfaces revealed features reminiscent of locally ductile and locally brittle failure mechanisms, namely, microscopic voids, dimples, SiC particulate cracking, and microscopic cracks. During cyclic deformation the large mismatch in strain carrying capability that exists between the hard, brittle and elastically deforming SiC particle and the soft, ductile and plastically deforming magnesium alloy metal-matrix results in a concentration of
stress at and near the reinforcing SiC particle making it a favorable site for failure by cracking (Fig. 11c). The void nucleation and cracking of the SiCp suggests that local plastic strains dominate in regions containing a high volume fraction of the SiC particulate reinforcement. Fracture of the reinforcing SiCp, in the plastic zone ahead of an advancing crack, generates conditions conducive for crack tip extension, which in turn is favorable for enhancing crack propagation rate. For a constant volume fraction (‘f’) of the SiCp reinforcement the tendency for it to fracture by cracking is greatest when the tensile stress exceeds the fracture strength of the particle. A gradual increase in the local stress concentration favors the initiation of microcracking at the SiC particles either individually dispersed or in clusters, while concurrently favoring decohesion or separation at their interfaces with the metal matrix. Numerous microscopic cracks were found at the reinforcing SiC particulates. Besides, the matrix between
220
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
Fig. 11. Scanning electron micrographs of the Z6/SiC/20p-F composite cyclically deformed at stress amplitude of 124 MPa, NfZ364,874 cycles, showing: (a) overall morphology, (b) high magnification of (a) showing population of microscopic voids, (c) cracked SiC particulate, (d) region of overload showing microscopic cracks and dimples.
particulate clusters is subjected to a high triaxial stress, which favors fracture of the matrix while concurrently exacerbating microscopic crack initiation at the SiC reinforcing particulates. The microscopic cracks grow through the SiC particulate clusters and link by fast fracture through the magnesium alloy metal matrix to adjacent SiC particles. This results in a macroscopic crack. Rapid propagation of the macroscopic cracks along with the microscopic cracks is responsible for the inferior cyclic stress resistance at ambient temperature (27 8C). Particulate fracture aided by decohesion at its interfaces with the metal matrix, void nucleation, growth and coalescence serves to increase the rate of microcrack growth by providing conditions conducive for crack-tip extension. For a constant volume fraction of the SiC particulate reinforcement, the probability for the reinforcing particle to fracture increases at the higher cyclic stress amplitudes and concomitant shorter fatigue life.
5. Conclusions A comprehensive study of the tensile deformation, cyclic stress-controlled fatigue and final fracture (quasi-staticC cyclic) behavior of Z6/SiCp metal-matrix composite provides the following highlights: 1. The initial microstructure of the Z6/SiC/20p-F metal matrix composite revealed a near-uniform distribution of the reinforcing SiC particulates in the longitudinal direction of the extruded composite plate. At intervals, an agglomeration or clustering of the reinforcing SiC particulates was observed. 2. The presence of hard, brittle and elastically deforming SiC particulates in a soft, ductile and plastically deforming magnesium alloy metal matrix favors the initiation of fine microscopic cracks at low values of applied stress. Extensive fractography of the failed
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
samples revealed limited ductility on a macroscopic scale, but microscopic features were reminiscent of locally ductile and brittle mechanisms. Fracture of the metal matrix between the dispersed SiC particulates and particulate clusters coupled with failure of the reinforcing SiC particulate by cracking facilitates rapid growth of the microscopic cracks and eventual coalesce by fracture through the ductile metal matrix. This results in macroscopic failure and low tensile ductility. 3. Influence of temperature on stress amplitude–fatigue life response revealed degradation in cyclic fatigue life with an increase in test temperature from 27 to 150 8C. When the cyclic stress amplitude is normalized with respect to yield strength (sYS) and ultimate tensile strength (sUTS) of the material at the test temperature, the variation with fatigue life reveals an improved cyclic stress resistance at 150 8C. The observed improvement in cyclic fatigue life is more pronounced at the higher ratios of stress than at the lower ratios of stress. 4. At a given temperature, cyclic fracture morphology was found to be essentially similar over the range of cyclic stress amplitudes. Macroscopic observations revealed the fracture surface to be brittle, i.e.: low cyclic ductility, at all cyclic stress amplitudes, but having microscopic features reminiscent of locally ductile and brittle mechanisms. Constraints in mechanical deformation, induced in the metal matrix, by the hard, brittle and elastically deforming SiCp reinforcement phase coupled with local stress concentration effects at the matrix-reinforcement particle interfaces promote failure through the conjoint influences of particle cracking and fast fracture through the matrix. Fracture of the plastically deforming matrix occurred through the formation, growth and coalescence of the voids.
Acknowledgements This research was jointly supported by State of Ohio: Board of Regents (Columbus, OH, USA) (Grant No. R4414OBR), the University of Akron (Akron, OH, USA) and personal funds of Dr Srivatsan. Sincere thanks and appreciation are extended to Dow Chemical Company (Freeport, TX, USA) for providing the material used in this study. Sincere thanks and abundance of appreciation is extended to the unknown reviewers whose comments, suggestions and corrections to text have most certainly helped strengthen the overall manuscript. One of the authors (Dr TSS) extends profound thanks to Mr G. Guruprasad (Graduate Student: Department of Mechanical Engineering) for terminal assistance with recompiling the manuscript for re-submission.
221
References [1] Lewandowski JJ. In: Clyne TW, editor. Metal matrix composites, vol. 3. Amsterdam: Elsevier; 2000. p. 151–87. [2] Divecha AP, Fishman SG, Karmarkar SD. Synthesis of metal-matrix composites. J Metals 1981;33:12–15. [3] Stephens JR. High temperature metal matrix composites for future aerospace systems. NASA TM 1987;100–212. [4] Nair SV, Tien JK, Bates RC. Int Metals Rev 1985;30(6):285–97. [5] Srivatsan TS, Sudarshan TS, editors. Rapid solidification technology: an engineering guide. Lancaster, PA: Technomic; 1993. p. 603–700. [6] Koczak MJ, Khatri SC, Allison JE, Bader MG. In: Suresh S, Mortensen A, Needleman A, editors. Fundamentals of metal matrix composites. Boston, MA: Butterworth/Heinemann; 1993. p. 297–317. [7] Taya M, Arsenault RJ. Metal matrix composites: thermomechanical behavior. Elmsford, NY: Pergamon Press; 1989. [8] McDanels DL. Metall Trans 1985;16A:1105–15. [9] Allison JE, Jones JW. Fatigue behavior of discontinuously-reinforced metal-matrix composites. In: Suresh S, Mortensen A, Needleman A, editors. Fundamentals of metal matrix composites. London: Heinemann; 1993. p. 269–96. [10] Srivatsan TS, Sudarshan TS, Lavernia EJ. Processing of discontinuously reinforced composite materials by rapid solidification. Prog Mater Sci: Int Rev J 1995;39(4/5):317–409. [11] Christman T, Suresh S. Microstructural development in an aluminum alloy-SiC whisker composite. Acta Metall 1988;36(7): 1699–704. [12] Llorca J, Needleman A, Suresh S. An analysis of the effects of matrix void growth on deformation and ductility in metal-ceramic composites. Acta Metall 1991;39(10):2317–35. [13] Christman T, Suresh S. Mater Sci Eng 1988;102:211–20. [14] Christman T, Needleman A, Suresh S. An experimental and numerical study of deformation in metal-ceramic composites. Acta Metall 1988; 37:211–20. [15] Hunt Jr. WH. Discontinuously-reinforced aluminum: the second generation. In: Ravi VA, Srivatsan TS, Moore JJ, editors. Processing and fabrication of advanced materials III. Warrendale, PA: TMS: The Minerals, Metals and Materials Society; 1994. p. 663–90. [16] Mikucki BA, Mercer II WE, Green WG. Extruded magnesium alloys reinforced with ceramic particles. Annual Meeting of ASM International; 1990. [17] Chaudhury PK, Rack HJ, Mikucki BA. J Mater Sci 1991;26: 2343–7. [18] Couling SL. Use of air/CO/SF mixtures for improved protection of molten magnesium. Paper presented at the 36th Annual World Conference on Magnesium, Oslo, Norway; 24–28 June 1979. p. 54. [19] ASTM Standard E-8. Standard Test Method for Tension Testing of Metallic Materials. American Society for Testing and Materials, Race Street, Philadelphia, PA, USA; 1993. [20] Shang JK, Ritchie RO. Crack bridging by uncracked ligaments during fatigue crack growth in silicon carbide reinforced aluminum alloy composite. Metall Trans 1989;20A:897–908. [21] Osman TM, Lewandowski JJ, Hunt Jr WH. Microstructure-property relationships for an aluminum/SiCp composite with different deformation histories. In Processing and fabrication of fiber-reinforced MMCs. Proceedings of an international conference, Montreal. Materials Park, OH, USA: American Society for Materials International; September 1990. p. 209–16. [22] Srivatsan TS, Mattingly J. Influence of heat treatment on the tensile behavior of an aluminum alloy-ceramic particle composite. J Mater Sci 1993;28:611–20. [23] Srivatsan TS. Influence of secondary processing on cyclic fracture behavior of an aluminum-ceramic particle composite. In: Ravi VA,
222
T.S. Srivatsan et al. / Composites: Part B 36 (2005) 209–222
Srivatsan TS, Moore JJ, editors. Processing and fabrication of advanced materials III. TMSD, Warrendale, PA: TMS: The Minerals, Metals and Materials Society; 1994. p. 699–731. [24] McClintock FA. Ductility. Materials Park, OH, USA: American Society for Metals; 1968 p. 256–66. [25] Van Stone RH, Cox TB, Low Jr JR, Psioda JA. The fracture behavior of metals. Int Metals Rev 1975;30:157–71. [26] Argon AS, Im J, Safoglu R. Cavity formation from inclusions in ductile fracture. Metall Trans 1975;6A:825–31.
[27] Lloyd DJ. Aspects of fracture in particulate-reinforced metal-matrix composites. Acta Metall Mater 1991;39(1):59–71. [28] Derby B, Mummery PM. In: Suresh S, Mortensen A, Needleman A, editors. Fundamentals of metal-matrix composites. Boston, MA: Butterworth/Heinemann; 1992. p. 251–71. [29] Lloyd DJ. Int Mater Rev 1994;39:1–23. [30] Haasen DF, Crowe CR, Ahearn JS, Cooke DC. In: Early JG, Shives TR, Smith JH, editors. Failure mechanisms in high performance materials, 1984. p. 147.