碳纤维复合材料英文文献

Journal of Materials Processing Technology168(2005)262–269Process optimisation for a squeeze cast magnesium alloy metalmatrix compositeM.S.Yong a,∗,A.J.Clegg ba Singapore Institute of Manufacturing Technology,71Nanyang Drive,Singapore638075,Singaporeb Wolfson School of Mechanical and Manufacturing Engineering,Loughborough University,Loughborough,Leicestershire LE113TU,UKReceived5January2004;received in revised form5January2004;accepted27January2005AbstractThe paper reports the influence of process variables on a zirconium-free(RZ5DF)magnesium alloy metal matrix composite(MMC) containing14vol.%Saffilfibres.The squeeze casting process was used to produce the composites and the process variables evaluated were applied pressure,from0.1MPa to120MPa,and preform temperature from250◦C to750◦C.The principalfindings from this research were that a minimum applied pressure of60MPa is necessary to eliminate porosity and that applied pressures greater than100MPa causefibre clustering and breakage.The optimum applied pressure was established to be80MPa.It was also established that to ensure successful preform infiltration a preform temperature of600◦C or above was necessary.For the optimum combination of a preform preheat temperature of600◦C and an applied pressure of80MPa,an UTS of259MPa was obtained for the composite.This represented an increase of30%compared to the UTS for the squeeze cast base alloy.©2005Elsevier B.V.All rights reserved.Keywords:Magnesium alloys;Squeeze casting;Metal matrix composites;Mechanical properties1.IntroductionMetal matrix composite(MMC)components can be man-ufactured by several methods.The metal casting route is espe-cially attractive in terms of its ability to produce complex near net shapes.However,castings produced by conventional cast-ing processes may contain gas and/or shrinkage porosity.The tendency for porosity formation will be exacerbated whenfi-bres are introduced because they tend to restrict theflow of molten metal and cause even greater gas entrapment within the casting.It is pointless to usefibres to reinforce a casting if defects are present,since the addition offibres will not com-pensate for poor metallurgical integrity.In order to fulfil the potential offibre reinforcement and produce pore free cast-ings the squeeze casting process can be selected.The unique feature of this process is that metal is pressurised throughout solidification.This prevents the formation of gas and shrink-age porosity and produces a metallurgically sound casting.∗Corresponding author.E-mail address:msyong@.sg(M.S.Yong).Selection of this process is also based on its suitability for mass production,ease of fabrication and its consistency in producing high quality composite parts.With the development of MMCs,magnesium alloys can better meet the various demands of diverse applications.The addition of reinforcement to magnesium alloy produces su-perior mechanical properties[1–3]and good thermal stability [4,5].Of the various composite types,the discontinuous and randomly orientedfibre-reinforced composites provide the best“value to strength ratio”.Despite the potential advantage of using magnesium MMC for lightweight and high strength applications,little is known about the influence of squeeze infiltration parame-ters.Key parameters,such as applied pressure and preform temperature must be optimised,especially for the squeeze infiltration of a magnesium–zinc MMC.These process pa-rameters were researched and the results are presented in this paper.However,it wasfirst necessary to select appropriate fibres and binders since their selection is fundamental to the success of the MMC.The main criterion determining the se-lection offibre type is compatibility with the matrix.Two0924-0136/$–see front matter©2005Elsevier B.V.All rights reserved. doi:10.1016/j.jmatprotec.2005.01.012M.S.Yong,A.J.Clegg/Journal of Materials Processing Technology168(2005)262–269263fibre types that are known to be compatible with magnesium are Saffil and carbon[6].Silica and alumina-based binders are widely used in preform production,mainly due to their high temperature properties[7].However,there are concerns about chemical reactions between magnesium alloys and sil-ica[8].To ensure full infiltration of liquid metal into thefibre pre-form,researchers[9–11]have emphasised the importance of preheating the preforms.However,there has been lit-tle research to determine optimum preform temperature for magnesium alloys and that reported has focused on AZ91 (magnesium–aluminium)alloy.The wetting capability of these alloys is different,for instance the wetting and the in-terfacial reaction between Al2O3reinforcement and cerium, lanthanum(both rare earth elements)or magnesium is far better in comparison to aluminium.Most work on applied pressure has focused on aluminium alloys and their composites.However,Ha[12]and Chadwick [13]investigated the influence of applied pressure on the short freezing range Mg–Al family of alloys.The effect on solid-ification will inevitably be different for long freezing range alloys,such as the Mg–Zn family alloys that are the focus of this research.The difference in solidification morphology will be significant when infiltrating the melt into a porousfi-bre preform.Long freezing range alloys retain a liquid phase over a longer period during infiltration and this may promote better infiltration,reduce voids and consequently improve the soundness of the composite.2.Experimental methodologyA zirconium-free magnesium–4.2%zinc–1%-rare earths alloy,designated RZ5DF,was used for this research.Several fibre preform materials,proportions and binder systems,were evaluated to determine their compatibility with the magne-sium alloys and the mechanical properties that they delivered to the composite[14].This preliminary research established that a compopsite based on a silica-bonded,14vol.%Saffil fibre preform delivered the best characteristics in terms of ease of production and maximum‘value to strength ratio’.The effect of applied pressure,between0.1MPa and 120MPa,on the RZ5DF-14vol.%Saffilfibre composite was first evaluated.The maximum permissible applied pressure was limited by both the capability of the squeeze casting press and die design.The metal pouring temperature was main-tained at750◦C,the die temperature at250◦C,the duration of applied pressure at25s,and delay before application of pressure at4s.These conditions replicated those employed for the base alloy that was reported previously[15].Following this,the influence of preform temperature was evaluated for a restricted range of applied pressures.Four preform temperatures were selected:250◦C(similar to the die temperature),400◦C(intermediate temperature),600◦C (at which temperature the RZ5DF alloy is a mixture of liquid and solid),and750◦C(at which temperature the RZ5DF alloy is in the fully molten state).These experiments were conducted at three applied pressures:60MPa,80MPa and 100MPa.The mechanical properties were evaluated using tensile and hardness tests.These tests were complemented by optical microscopy and,for the tensile fracture surfaces,SEM.2.1.Test castingThe test casting was a rectangular plate of126mm in length,75mm in width and16mm in depth.2.2.Melt processingThe alloy was melted in an electric resistance furnace us-ing a steel crucible,thefluxless method and an argon gas cover.The die was coated with boron nitride suspended in water to protect it from excessive wear.2.3.Tensile testingTensile tests were conducted on a50kN Mayes testing ma-chine using position control.Modified test specimens were machined according to BS18(1987)and magnesium Elek-tron Ltd RB4specifications[16].2.4.Hardness testingHardness was measured to determine and study the influ-ence of reinforcement on the magnesium and the isotropy of fibre distribution.The locations of hardness measurements are shown in Fig.1.Hardness measurements were conducted using the Rockwell B scale for both the alloys and com-posites.The preference for the Rockwell rather than Vick-ers hardness measurement was due to the larger indentation needed to ensure a more consistent measurement on the com-posite.The area of the Vickers hardness indentation is so small that,in some cases,the measurement could be taken from the hardfibre causing large variations in hardness val-ues.Fig.1.Locations of hardness measurements(each dot represents the position of a hardness measurement)taken in both‘Longitudinal’and‘Transverse’directions.264M.S.Yong,A.J.Clegg /Journal of Materials Processing Technology 168(2005)262–2692.5.MetallographyAn optical microscope and stereoscan 360electrom mi-croscope (SEM)were used to examine the microstructure of the MMC specimens.Metallographic samples were prepared using standard techniques and were etched using an acetic pi-cral solution.The electron microscope was equipped with a back-scatter detector and was used to characterise fracture surfaces from the tensile test specimens.2.6.Cell sizeThe cell size was established using the intersection method.Five areas were selected at random and 21mea-surements of cell size were taken for each area.The average value for the 105readings was determined.3.Results and observationsThe results are reported in the sequence in which the ex-periments were conducted.In the first series of experiments,the effect of applied pressure was evaluated.In the second se-ries,the combined influences of applied pressure and preform preheat temperature were evaluated.3.1.Series 1experiments:the influence of applied pressure3.1.1.Tensile propertiesThe effect of applied pressure on UTS and ductility of squeeze cast,RZ5DF-14vol.%,Saffil fibre composites is shown in Fig.2.It can be seen that the highest UTS value was obtained with an applied pressure of 80MPa.It would appear from the figure that a pressure in excess of 40MPa is essential to develop a significant improvement in UTS but that levels above 80MPa have a detrimental effect.3.1.2.HardnessThe hardness values along the longitudinal and transverse directions of the composite castings produced atdifferentFig.2.The effects of squeeze infiltration applied pressure on the tensile properties of the RZ5DF matrix with 14vol.%fraction Saffilfibres.Fig.3.The average material hardness along the longitudinal and transverse direction of the squeeze infiltrated RZ5DF alloy with 14vol.%fraction Saffil fibres,cast with constant pouring temperature of 750◦C and die temperature of 250◦C.applied pressures are shown graphically in Fig.3.Whilst the dominating influence on hardness is provided by the presence of the Saffil fibres,the results show that the hardness at the two lowest levels of applied pressure (0.1MPa and 20MPa)is distinctly lower than that associated with applied pressure levels of 40MPa and above.3.1.3.MetallographyMetallography was conducted to examine the influence of applied pressure on the cast structure.Selected opti-cal microstructures are presented in Fig.4.The metal-lographic examination identified the presence of microp-orosity in those samples produced with applied pressures below 60MPa.The microporosity,as expected,occurred mainly at cell boundaries and was most easily confirmed by adjusting the depth of field.It also identified the ten-dency for fibre clustering and fracture at applied pressures greater than 80MPa.The presence of fractured fibres is demonstrated more clearly in the SEM micrographs shown in Fig.5.These micrographs show fractured fibres in the plane transverse to that of load application during the tensile test.3.2.Series 2experiments:the influence of preform temperatureThe preliminary experiments showed that the optimum applied pressure was 80MPa.However,to ensure robustness in the experimentation,the effects of preform preheat temper-ature were evaluated for the optimum applied pressure and pressures of 60MPa and 100MPa.3.2.1.Tensile testsThe effects of preform temperature and applied pressure on UTS are summarised in Fig.6.The results show that a preform preheat temperature of 750◦C produced the most consistent UTS values across the range of applied pressuresM.S.Yong,A.J.Clegg/Journal of Materials Processing Technology168(2005)262–269265Fig.4.Optical microstructure of squeeze infiltrated RZ5DF-14vol.%fraction Saffilfibres produced under(i)atmospheric pressure,0.1MPa,applied pressure of(ii)20MPa,(iii)40MPa,(iv)60MPa,(v)80MPa,(vi)100MPa and(vii)120MPa.and that the maximum UTS of259MPa was obtained with a preform temperature of600◦C and an applied pressure of 80MPa.These results confirm the status of80MPa as the optimum value of applied pressure.3.2.2.HardnessThe results of the hardness tests are shown in Fig.7.The greatest variation in hardness was demonstrated by the test casting produced with the lowest value of appliedpressure Fig.5.SEM micrograph of the fracture face of a squeeze infiltrated RZ5DF-14vol.%fraction Saffilfibres produced under applied pressure of(i)100MPa and (ii)120MPa.266M.S.Yong,A.J.Clegg/Journal of Materials Processing Technology168(2005)262–269Fig.6.The plot of UTS for RZ5DF-14vol.%Saffil MMC produced from various combinations of applied pressure and preform temperature. (60MPa)and preform temperature of400◦C.The range of variation was±8HRB compared to±6HRB observed for the other combinations of preform temperature and applied pressure.3.2.3.MetallographyMetallographic examination of the composite structures showed that more densely packedfibres occurred at the pre-form surface at the lowest preform temperature.This effect is illustrated in Fig.8.The sequence of microstructures show that preform deformation andfibre clustering were less evi-dent at higher preform temperatures.The SEM micrographs of tensile fracture surfaces,Fig.9,confirm the clustering of fibres and provide evidence offibre tofibre contact,for the preheat temperature of400◦C.This effect was not evident for the preheat temperature of600◦C.4.DiscussionTo achieve the successful infiltration of afibre preform the liquid metal must penetrate the preform completely.Potential barriers to this are presented by:the density of the preform, which can be represented by the preform permeability[14]; Fig.7.The average material hardness along the longitudinal and transverse direction of the squeeze infiltrated RZ5DF alloy with14vol.%fraction Saffilfibres produced under different combinations of preform temperatures and appliedpressures.Fig.8.A micrograph taken at the preform infiltration region of a squeeze infiltrated specimen produced with a preform temperature of(i)750◦C,(ii)600◦C, (iii)400◦C and(iv)250◦C.M.S.Yong,A.J.Clegg/Journal of Materials Processing Technology168(2005)262–269267an insufficient pressure head,necessary to displace the air and overcome resistances to metalflow;and/or a low pre-form temperature that promotes premature solidification of the solid before complete infiltration.Increasing either the applied pressure or the preform pre-heat temperature,independently or in combination,may im-prove infiltration.However,there may be adverse conse-quences.Too high a level of applied pressure may physi-cally damage the preform through compression.This leads to compacted preforms that resist infiltration together withfi-bre clustering andfibre breakage that reduce thefibres’effec-tiveness for strengthening the matrix.Although researchers [11,17,18]have resorted to high preform temperatures to achieve infiltration,this too can have adverse effects.For example,an increased heat content in the system may retard solidification.This in turn extends the time during which there is the opportunity for adverse interfacial reactions to occur between the alloy andfibres.Furthermore,an extended pe-riod of solidification can promote the formation of larger cell sizes that in turn impair the mechanical properties.The influence of applied pressure is quite clearly demon-strated in Fig.2.Thefigure can be divided into three distinct regions:<60MPa,61–90MPa,>91MPa.Thefirst of these regions is associated with the presence of porosity and voids in the castings and this porosity is associated with low UTS values.As the applied pressure is increased the porosity is eliminated and the composite develops its optimum UTS of 259MPa at an applied pressure of80MPa.Thereafter,an increase in applied pressure producesfibre clustering and breakage leading to more initiation points for fracture and so the UTS declines.The tensile evidence is supported by evi-dence from hardness tests and metallography.The presence of porosity,revealed in Fig.4,adversely affects the hardness of the castings.Quite simply,low levels of applied pressure are not sufficient to either suppress porosity formation or com-pletely infiltrate thefibre preform.It is interesting to note that the optimum applied pressure level of80MPa for the compos-ite is20MPa higher than that necessary to develop the highest level of strength in thefibre-free base alloy[15].Metallo-graphic examination revealed that preform deformation and fibre clustering was less evident andfibres were less densely packed at the surfaces of the preforms preheated to600◦C or 750◦C(see Fig.8)when compared with400◦C and250◦C.It was found that the highest preform temperature(750◦C) produced the most consistent UTS values over the range of applied pressures considered.This preform temperature is above the liquidus temperature of the alloy.It would,there-fore,be expected that infiltration of the preform would not be impeded by the early onset of solidification of the alloy on thefibre preforms.The preform temperature of600◦C pro-duced a higher variation in UTS than was observed for the 750◦C preform preheat temperature.However,the highest value of UTS of all the experiments was produced with this preheat temperature in combination with an applied pressure of80MPa.A preform temperature of750◦C supported a wider range of applied pressure because,even at the lowest level of 60MPa,there was a minimal resistance to infiltration.It was also noted that there was less variation infibre distri-bution.An even distribution offibres was also evident in the specimens produced at a preheat temperature of600◦C, see Fig.9.This temperature is33◦C below the alloy’s liq-uidus temperature.Although infiltration was not a problem, it can be postulated that solidification would occur quite quickly under these conditions.This postulation is supported by microstructural evidence and cell size measurements,see Fig.8,that show a smaller cell size,associated with better UTS,in the samples produced with a preform temperature of 600◦C.With preheat temperatures of400◦C and,especially, 250◦C the UTS values are generally poor and there is clear evidence of ineffective infiltration.The microstructural evi-dence clearly shows that preform deformation,fibre cluster-ing andfibre breakage is evident to varying degrees.However, such effects were not uniform and produced inconsistent ef-fects.For example,for the combination of400◦C and60MPa applied pressure,microstructural evaluation,see Fig.10,re-vealed a high concentration offibres in the centre of the infil-trated preform.This effect was caused by two factors.Firstly, the low preform preheat temperature promoted rapid solid-ification of the alloy prior to the application of pressure at both the preform surface and at locations near to the die wall. Secondly,the low applied pressure resulted in irregular and curtailed infiltration.In consequence,the applied pressure compacts rather than infiltrates the preform.This produces in-filtrated regions that have a higher concentration offibresand Fig.9.SEM micrograph of the fracture face of a squeeze infiltrated RZ5DF-14vol.%Saffil MMC produced with(i)400◦C and(ii)600◦C preform temperature.268M.S.Yong,A.J.Clegg /Journal of Materials Processing Technology 168(2005)262–269Fig.10.Microstructure showing different parts of the squeeze infiltrated RZ5DF-14vol.%fraction Saffil specimen produced with a preform temperature of 400◦C and applied pressure of 60MPa.The sequence is (i)top,(ii)centre and (iii)bottom portion of the fabricated composite.this can produce higher values of UTS,see Fig.10.However,the effect is inconsistent and therefore undesirable.Hardness measurements also confirmed the inconsistency.For exam-ple,the specimen produced with 60MPa and 400◦C preheat demonstrated the greatest variation in hardness,see Fig.7,and this was attributable to the central clustering of fibres.Ex-amination of the fracture surface of the specimen produced with a preform temperature of 400◦C and an applied pres-sure of 80MPa clearly shows the fibres in contact with one another,see Fig.9.4.1.The influence of zincAlloying magnesium with 4.2%of the lower melting point metal zinc produces a binary alloy that has a long freezing range.Experimentation [16]determined the values of the liquidus and solidus of the RZ5DF alloy to be 633◦C and 474◦C,respectively,a freezing range of 159◦C.Although long freezing range alloys are the most prone to shrinkage porosity,this problem is overcome by squeeze casting.The long freezing range may in fact be beneficial in the produc-tion of a composite since the extended period during which a liquid phase is present may promote infiltration.The pres-ence of zinc may also be significant for the preform preheat temperature.The results show that the optimum UTS of 259MPa was obtained with a preform temperature of 600◦C,a tempera-ture just 33◦C below the alloy’s liquidus temperature.Cell size measurements revealed that specimens produced at this preheat temperature had a smaller average cell size,typically 30␮m,see Fig.8.For specimens produced with preheat tem-peratures of 750◦C,400◦C and 250◦C the average cell size was >50␮m.This variation can be explained by consider-ation of the nucleation and growth sequence in the various specimens.The high preform preheat temperature retards the rate of solidification because time is necessary for the heat of the preform to be transferred through the alloy to the die.Nucleating cells have time to grow.Conversely,at low pre-heat temperatures of 400◦C and 250◦C,the alloy solidifiesquickly in contact with the relatively cold fibres.The first solid formed is rich in the primary phase and the remaining liquid becomes richer in the low melting point eutectic.Al-though primary phase still forms by nucleation and growth in the inter-fibre regions,the number of cells formed is reduced and their size is larger.5.Conclusions1.The optimum applied pressure for the squeeze casting of RZ5DF-14vol.%Saffil fibre composites was determined to be 80MPa.At applied pressures below 60MPa,micro-porosity was not suppressed.Conversely,a high applied pressure of 100MPa or above causes fibre clustering and breakage and a concomitant reduction in UTS.2.The optimum preform preheat temperature was estab-lished to be 600◦C.At this temperature consistent fibre in-filtration was achieved and the optimum cell size of 30␮m was obtained in the matrix.3.The optimum combination of applied pressure and pre-form preheat temperature was determined to be 80MPa and 600◦C,respectively.For this combination,a UTS value of 259MPa was obtained.The composite delivered a 30%increase in UTS compared with that developed in the squeeze cast base alloy.AcknowledgementsDr.Yong gratefully acknowledges the receipt of an Over-seas Research Students Award and a Loughborough Univer-sity Research Studentship.References[1]K.Purazrang,P.Abachi,K.U.Kainer,Investigation of the mechan-ical behaviour of magnesium composites,Composites 25(4)(1994)296–302.M.S.Yong,A.J.Clegg/Journal of Materials Processing Technology168(2005)262–269269[2]K.Purazrang,P.Abachi,K.U.Kainer,Mechanical behaviour of mag-nesium alloy MMCs produced by squeeze casting and powder met-allurgical techniques,Compos.Eng.3(6)(1993)489–505.[3]O.Ottinger,G.Grau,R.Winter,R.F.Singer,The effect of alu-minium additions on the interfacial microstructure and mechanical properties of C/Mg composites,in:Proceedings of the10th Inter-national Conference on Composite Materials(ICCM10),vol.VI, Vancouver,Canada,August1995,pp.447–454.[4]A Materials Edge Report,Metal matrix composites in the automotiveindustry,Met.Bull.plc.,(1993)1–33.[5]W.Toaz,R.R.Bowles,D.L.Mancini,Squeeze casting compositecomponents for diesel engines,Ind.Heat.54(3)(1987)17–19. [6]P.K.Rohatgi,Advances in cast mmc,Adv.Mater.Process137(2)(1990)39–44.[7]T.W.Clyne,P.J.Withers,An Introduction to Metal Matrix Compos-ites,Cambridge University Press,Cambridge,1993.[8]B.Inem,G.Pollard,Interface structure and fractography of amagnesium-alloy metal matrix composite reinforced with SiC parti-cles,J.Mater.Sci.28(1993)4427–4434.[9]H.Fukunaga,K.Goda,Fabrication offiber reinforced metal bysqueeze casting,Bull.JSME27(228)(1984)1245–1250.[10]H.Fukunaga,Processing aspects of squeeze casting for shortfi-bre reinforced metal matrix composite castings,Adv.Mater.Manuf.Process3(4)(1988)669–687.[11]J.Kiehn,W.Riehemann,K.U.Kainer,P.V ostry,I.Stulikova,B.Smola,Annealing effects in shortfibre reinforced and unreinforcedMg–Ag–Nd–Zr alloy,in:Proceedings of the Third International Magnesium Conference,Manchester,UK,April,1996,pp.663–676.[12]T.U.Ha,Squeeze casting of magnesium-based alloys and their metalmatrix composites,Ph.D.Thesis,University of Southampton(1988).[13]G.A.Chadwick,Squeeze casting of magnesium alloys andmagnesium-based metal matrix composites,in:Proceedings of Mag-nesium Technology,London,The Institute of Metals,November 1986,pp.75–82.[14]M.S.Yong,R.I.Temple,A.J.Clegg,Influence offibre preform per-meability on infiltration of magnesium–zinc base alloys,in:Pro-ceedings of Magnesium Alloys and their Applications,Wolfsburg, Germany,18–20November,2003,pp.348–353.[15]M.S.Yong,A.J.Clegg,Process optimisation for a squeeze cast mag-nesium alloy,J.Mater.Process.Technol.145(January(1))(2004) 134–141.[16]M.S.Yong,Process optimisation of squeeze cast magnesium–zinc–rare earth alloys and shortfibre composites,Ph.D Thesis, Loughborough University,1999.[17]S.Kamado,Y.Kojima,Microstructure and tensile properties ofMg–Zn based alloy composite reinforced with Al2O3shortfibre and9Al2O3·2B2O3whisker,p.Mater6(3)(1997) 159–167.[18]J.H.Hsieh,C.G.Chao,Effect of magnesium on mechanical prop-erties of Al2O3/Al–Zn–Mg–Cu metal matrix composites formed by squeeze casting,Mater.Sci.Eng.A.214(1996)133–138.。