The repeated thermal shock behaviors of a ZrB2

The repeated thermal shock behaviors of a ZrB 2–SiC composite heated by electric resistance methodSonghe Meng,Fei Qi ⁎,Hongbo Chen,Zhi Wang,Guanghui BaiCenter for Composite Materials,Harbin Institute of Technology,Harbin 150001,Chinaa b s t r a c ta r t i c l e i n f o Article history:Received 15April 2010Accepted 24June 2010Keywords:ZrB 2-based ceramic MicrostuctureRepeated thermal shocks Mechanical propertiesRepeated thermal shocks of ZrB 2–20vol.%SiC (ZrB 2–SiC)composite was investigated by the electric resistance method with 330A and 4V.It was found that the maximum temperature of the specimen center was obtained in 20s and the specimen was cooled down to room temperature in 10s due to a coolant system,and then the specimen was heated again.The thermal shock treatment increased the flexural strength with a peak at 20cycles.Even though 30cycles showed the minimum strength,it was still higher than that of the unshocked specimen.The increase in the strength was attributed to the formation of the oxide layer and the compression stress that resulted from volume expansion upon conversion of ZrB 2and SiC,to ZrO 2,B 2O 3and SiO 2.©2010Elsevier Ltd.All rights reserved.1.IntroductionZirconium diboride (ZrB 2)based ultrahigh temperature ceramic matrix composites are a leading candidate as potential materials for a variety of high temperature structural applications,such as thermal protection structures for leading edge parts on hypersonic reentry space vehicles at over 1800°C [1–3].It is primarily because ZrB 2has an excellent combination of higher melting point (3245°C)[1],high strength [2],good thermal stability [3],corrosion resistance [4],physical [5],chemical [6]and oxidation resistance properties [7].Up to now,it is known that the addition of appropriate amounts of SiC to these materials improves the oxidation resistance and mechanical properties of ZrB 2based ceramics [8,9].Although these materials have many advantages,their intrinsic characteristics such as low fracture toughness (premature failure due to brittle fracture),poor thermal shock resistance are still obstacles for them to be used widely,especially for applications with high heat transfer and/or rapid environmental temperature changes,such as thermal protection structures for leading edge parts on hypersonic reentry space vehicles and hypersonic aircraft [10].Aerodynamic heating increases as the cube of speed V,but aerodynamic drag increases only as the square of speed V.The ascent peak stagnation point and wing leading-edge equilibrium wall temperatures of hypersonic trans-atmospheric vehicles are about 3000K,respectively [11].ZrB 2–20vol.%SiC (ZrB 2–SiC)composites,because of inherently low fracture toughness,are susceptible to catastrophic fracture caused by the thermal stresses under the thermal shock conditions.There are many methods available for measurement of the thermal shock resistance of ceramic matrix composites.They can be broadly classi fied into two groups:(1)thermal shock during heating (ascendant)and (2)thermal shock during cooling (descendant)[12].Among these methods,the water quenching is the most popular one in which rectangular bars are heated to a desired temperature and then quenched into water bath of room temperature.The residual strength after the quench and the critical temperature difference (ΔT c)are the two parameters used to grade the thermal shock resistance of various materials.However,water quench experiment generates higher rates of heat transfer than a material would experience during its actual use [13].Therefore,the material fails relatively at a lower critical temperature difference.In addition,heat transfer coef ficient involved in a water quench method is dif ficult to measure due to vaporisation of water.Moreover,this method is of little relevance for ceramic materials used in aerospace,as thermal protection components where thermal shock occurs mostly during heating [14].Among ascending thermal shock tests,except for laser heating method,electric resistance heating method is cheap and rapid,which is easy to generate high temperatures suitable for thermal shock test.In the present work,the cycle heating and cooling was carried out by electric resistance heating method in order to investigate the repeated thermal shock resistance of ZrB 2–SiC composite.The purpose of this paper is to report the repeated thermal shock method and the repeated thermal shock resistance of ZrB 2–SiC composite.Furthermore,the microstructure of ZrB 2–SiC composite after the repeated thermal shock was investigated and discussed.Int.Journal of Refractory Metals and Hard Materials 29(2011)44–48⁎Corresponding author.Tel./fax:+8645186402432.E-mail address:hitqifei@ (F.Qi).0263-4368/$–see front matter ©2010Elsevier Ltd.All rights reserved.doi:10.1016/j.ijrmhm.2010.06.012Contents lists available at ScienceDirectInt.Journal of Refractory Metals and Hard Materialsj o ur n a l h o m e p a ge :ww w.e l s ev i e r.c o m/l o c a t e /I J R M H M2.ExperimentalCommercially available ZrB 2(Northwest Institute for Non-ferrous Metal Research,China)with a reported purity of N 99.5%(metal basis)and an averaged particle size of 5μm was used.The SiC powders (Weifang Kaihua Micro-powder Co.,Ltd.,China)had a reported purity of 99.5%and an average particle size of 2μm.Batch containing 80vol.%ZrB 2and 20vol.%SiC was milled in a polymer-coated bucket charged with ethanol using ZrO 2balls for 12h and dried in a rotating led powders were hot pressed at 1900°C for 1h under a uniaxial load of 30MPa in air atmosphere.The specimens with the tensile surface perpendicular to the hot pressing direction were cut from the same billet,and then polished to 2μm using diamond abrasives.The flexural strength of the specimens before and after the repeated thermal shocks was measured in a three-point bending on 3mm by 4mm by 36mm bars,using a 30mm span and a crosshead speed of 0.5mm min −1.The edges of all the specimens were chamfered to minimize the effect of stress concentration due to machining flaws.A minimum number of six specimens were measured for per experimental condition.The microstructure of the specimen was examined using scanning electron microscopy (SEM;FEI Sirion,Holland)along with energy dispersive spectroscopy (EDS,USA)for chemical analysis.The phase composition was determined by X-ray diffraction (XRD;Rigaku,Dmax-rb,CuKa=1.5418Å).A home-made high current electrical source with power of 50kW is shown in Fig.1,which was used for the repeated thermal shock test of specimens.The specimen was fixed between Cu electrode and the temperature of the specimen center was measured by electric thermo-couple with measurement range of 20–1000°C and multi-wavelength pyrometer with measurement range of 1000–1700°C.The voltage for the two ends of the Cu electrode was maintained at 4V and the current was increased linearly to 330A by program control.3.Results and discussionsFig.2A shows the change curve in temperature of the specimen center with increasing testing time.The maximum temperature of the specimen center was obtained in 20s and the specimen was cooled down to room temperature in 10s due to the coolant system,and then the specimen was again heated.One cycle took about 30s.During the heating period,the macro-scale temperature field is not uniform at the two ends and the middle of the specimen:the temperature of the middle part of the specimen was higher than the ends.This phenomena can be observed in many metallic specimen heated by electrical resistance method [15].The macrograph of the specimen before and after different thermal shocks is shown in Fig.2B.The surface of the specimen was oxidized to oxides when the specimenwas exposed to air of high temperature [16].The surface of the specimen darkened gradually from the center to the ends,which was attributed to the presence of the different oxides because of inhomogeneous temperature field.The oxidation area on the surface of the specimen increased as the thermal shock times increased.The ZrB 2–SiC composite billets used to produce specimen bars for the thermal shock studies had a measured bulk density of 5.50g/cm ing a rule of mixture calculation,and assuming that the true density was 6.09g/cm 3for ZrB 2and 3.21g/cm 3for SiC [17],the theoretical density of the ZrB 2–20vol.%SiC composite was calculated to be 5.51g/cm 3.Based on this true density,the ZrB 2–SiC composite billets had relative density of N 99.8%.Consequently,the porosity within the specimen was not considered to have a signi ficant effect on the repeated thermal shock behavior.Fig.3shows the micrographs of polished surface (A)and fracture surface (B)of the unshocked ZrB 2–SiC composite.It could be found that the darker phase was SiC and it appeared to be uniformly dispersed in the lighter ZrB 2matrix.No porosity within the fracture surface of the specimen was observed,which con firmed that relative density was credible.Some micro flaws such as pores and microcracks that formed during machine process were observed on the surface of the specimen.The micro flaws were detrimental to the flexural strength of the specimen because ZrB 2-based ceramics are brittle and sensitive to super ficial defects such as cracks and pores which can lead to unexpected catastrophic failure of ceramic matrix composites under lower stress load [18].Micrographs of the surface of ZrB 2–SiC composite after the different repeated thermal shocks at maximum temperature of 1600°C are shown in Fig.4.The surface of the ZrB 2–SiC composite was oxidized to oxides when the ZrB 2–SiC composite was exposed to air at high temperature,for example,ZrB 2is oxidized to ZrO 2and B 2O3Fig.1.Macrograph of home-made high current electricalsource.Fig.2.The change curve (A)in temperature of the specimen center with increasing testing time and macrograph (B)of the specimen thermal shocked for 10,20,30and 50cycles.45S.Meng et al./Int.Journal of Refractory Metals and Hard Materials 29(2011)44–48at above 650°C [19];SiC is oxidized to SiO 2and CO 2at above 800°C [16].The expected main reactions describing the oxidation process were shown as following:ZrB 2þ5=2O 2ðg Þ→ZrO 2ðs ÞþB 2O 3ðl Þð1ÞSiC þ3=2O 2→SiO 2ðl ÞþCO ðg Þð2ÞB 2O 3ðl Þ→B 2O 3ðg Þð3ÞB 2O 3ðl ÞþSiO 2ðl Þ→SiO 2:B 2O 3ðl Þð4ÞThe combination of EDS and XRD analysis (not shown here)con firmed that the white particles were ZrO 2phase because the white particle composed only of oxygen and zirconium elements,and a continuous glassy phase was SiO 2-rich glass that was expected to contain some B 2O 3during transient heating to 1600°C based on either incomplete evaporation of the B 2O 3by Reaction (3)or the continued production of B 2O 3beneath the outer scale by Reaction (1)[19].The surface of the ZrB 2–SiC composite was dimmed gradually as the repeated thermal shock cycles increased,which indicated that the SiO 2-rich glass gradually increased on the surface of the ZrB 2–SiC composite.The white ZrO 2particles were covered with a continuous SiO 2-rich glass.The formation of such structure was presumably attributed to volume expansion and/or the mutual wetting behavior of the ZrO 2and SiO 2-rich glass phases [16].The pits on the surface of the specimen were covered with SiO 2-rich glass phases.In addition,the number and size of the pits increased with increasing thermal shockcycles.Fig.3.Micrographs of polished surface (A)and fracture surface (B)of the unshocked ZrB 2–SiCcomposite.Fig.4.Micrographs of the specimen shocked for 10(A),20(B),30(C)and 50(D)cycles,the inserts were magni fication.46S.Meng et al./Int.Journal of Refractory Metals and Hard Materials 29(2011)44–48Fig.5shows the micrographs of fracture surface of the ZrB 2–SiC composite after thermal shocks at 1600°C for 10(A),20(B),30(C)and 50(D)cycles.After exposure in air at 1600°C for thermal shock of 10,20,30and 50cycles,the oxide scales were about 26,32,34and 49μm,respectively.The thickness of the oxide layer increased with the increasing thermal shocks.The analysis of the SEM observations combined with EDS (not shown here)demonstrated that,beginning from the surface,the oxide layer was composed of a continuous SiO 2-rich glassy phase,which is followed by a layer of ZrO 2and/or ZrB 2from which SiC had been partially depleted [20].Many clavate ZrO 2grains could be observed in the borosilicate layer.Such growth of zirconia during the heating process has been also reported in other work [21].In contrast to the microstructures reported in literatures [19,20],many pores in the oxide layer were readily observed and the appearance of the ZrO 2grains was not very similar to those reported in literature.An unaffected ZrB 2–SiC was followed by the SiC depleted region.The very clear interface between the oxide layer and unaffected matrix was easily observed,which was different to the microstructures reported in literature [20].Such discrepancy was presumably attributed to different heating methods,and further analysis and discussion would be investigated in future.This layered microstructure was similar to the structure reported for ZrB 2–SiC exposed to air in previous studies [22].Fig.6shows the flexural strengths of the specimen before and after thermal shocks which were plotted as a function of the thermal pared to the original strength of 440±47MPa,the strength of 514±34MPa for a thermal shock of 10cycles was obviously increased due to the compression stress and the formation of the oxide layer that healed the flaws on the surface of the specimen.As thermal shock cycle increased to 20times,the strength as high as 645±60MPa was signi ficantly improved,which was presumably attributed to the formation of the oxide layer and the compressionstress that resulted from volume expansion upon conversion of ZrB 2and SiC,to ZrO 2,B 2O 3and SiO 2[16].Furthermore,the escape of gas products such as CO and B 2O 3through the oxide layer resulted in the formation of the pores,which led to the high error and the decrease in the strength,especially for thermal shock cycle of 30times (494±147MPa).For thermal shock cycle of 50times,the strength was again recovered to 589±72MPa compared with that of the specimen shocked for 30times,which was attributed to the increase in the oxides.At the same time,the pores in the oxide layer were not increased.The specimens shocked for 30times showed a minimum strength,which was still higher than that of the unshocked specimen.The increase in the strength was attributed to the formation of the oxidelayer.Fig.5.Micrographs of fracture surface of the ZrB2–SiC composite after thermal shocks at 1600°C for 10(A),20(B),30(C)and 50(D)cycles.The thicknesses of oxide layer were 26,32,34,and 49μm,respectively.The whole scales were mainly divided into three-layers:a continuous SiO 2-rich glassy phase ①,ZrO 2and/or ZrB 2from which SiC had been partially depleted ②,and unaffected ZrB 2–SiC ③.Fig.6.Effect of increasing number of shock cycles on the flexural strength.47S.Meng et al./Int.Journal of Refractory Metals and Hard Materials 29(2011)44–484.ConclusionsRepeated thermal shocks for the ZrB2–SiC composite was tested by electrical resistance heating method.Theflexural strength of the specimen increased after repeated thermal shocks.The increase in the strength was presumably attributed to the continuous oxide layer that healed theflaws on the surface of the specimen.However,the strength was not increased monotonically,which had a reduction in the strength for thermal shock of30cycles.The escape of gas products such as CO and B2O3through the oxide layer resulted in the formation of the pores,which led to the decrease in the strength,especially for thermal shock cycle of30times.The specimens shocked for30times showed a minimum strength,which was still higher than that of the unshocked specimen.The increase in the strength was attributed to the formation of the oxide layer.References[1]Zhao Y,Wang LJ,Zhang GJ,Jiang W,Chen LD.Effect of holding time and pressureon properties of ZrB2–SiC composite fabricated by the spark plasma sintering reactive synthesis method.Int J Refract Met Hard Mater2009;27:177–80.[2]Chamberlain AL,Fahrenholtz WG,Hilmas GE,Ellerby DT.High-strength zirconiumdiboride-based ceramics.J Am Ceram Soc2004;87:1170–2.[3]Meng SH,Liu GQ,An J,Sun SL.Effects of different additives on microstructure andcrack resistance for an ultra-high temperature ceramic.Int J Refract Met Hard Mater2009;27:813–6.[4]Weng L,Zhang XH,Han WB,Han JC.Fabrication and evaluation on thermalstability of hafnium diboride matrix composite at severe oxidation condition.Int J Refract Met Hard Mater2009;27:711–7.[5]Weng L,Han WB,Li XY,Hong CQ.High temperature thermo-physical propertiesand thermal shock behavior of metal–diborides–based composites.Int J Refract Met Hard Mater2010;28:459–65.[6]Sun X,Han WB,Hu P,Wang Z,Zhang XH.Microstructure and mechanicalproperties of ZrB2–Nb composite.Int J Refract Met Hard Mater2010;28:472–4.[7]Wang HL,Chen DL,Wang CA,Zhang R,Fang DN.Preparation and characterizationof high-toughness ZrB2/Mo composites by hot-pressing process.Int J Refract Met Hard Mater2009;27:1024–6.[8]Chen DJ,Xu L,Zhang XH,Ma BX,Hu P.Preparation of ZrB2based hybrid compositesreinforced with SiC whiskers and SiC particles by hot-pressing.Int J Refract Met Hard Mater2009;27:792–5.[9]Zhu SM,Fahrenholtz WG,Hilmas GE.Influence of silicon carbide particle size onthe microstructure and mechanical properties of zirconium diboride-silicon carbide ceramics.J Eur Ceram Soc2007;27:2077–83.[10]Liu GQ,Meng SH,An J,Sun SL,Ding XH.The evaluation of reliability for zirconiumdiboride composite with additives of graphite.Int J Refract Met Hard Mater 2009;27:868–71.[11]Mahulikar SP.Theoretical aerothermal concepts for configuration design ofhypersonic vehicles.Aerosp Sci Technol2005;9:681–5.[12]Panda PK,Kannan TS,Dubois J,Olagnon C,Fantozzi G.Thermal shock and thermalfatigue study of ceramic materials on a newly developed ascending thermal shock test equipment.Sci Technol Adv Mater2002;3:327–34.[13]Zimmermann JW,Hilmas GE.Fahrenholtz WG.Thermal shock resistance of ZrB2and ZrB2–30%SiC.Mater Chem Phys2008;112:140–5.[14]Zhao SY,Zhang BM,Du SY.Probabilistic modeling of transient heat transfer andassessment of thermal reliability offibrous insulation under aerodynamic heating conditions.Int J Therm Sci2009;48:1302–10.[15]Chen HB,Wang Z,Meng SH,Bai GH,Qu W.The failure mechanism of ZrB2–SiC–graphite composite heated by high electric current.Mater Lett2009;63:2346–8.[16]Rezaie A,Fahrenholtz WG,Hilmas GE.Evolution of structure during the oxidationof zirconium diboride-silicon carbide in air up to1500°C.J Eur Ceram Soc 2007;27:2495-01.[17]Zhou SB,Wang Z,Zhang W.Effect of graphiteflake orientation on microstructureand mechanical properties of ZrB2–SiC–graphite composite.J Alloys Compd 2009;485:181–5.[18]Zhang XH,Xu L,Du SY,Han WB,Han JC.Crack-healing behavior of zirconiumdiboride composite reinforced with silicon carbide whiskers.Scr Mater2008;59: 1222–5.[19]Guo WM,Zhang GJ,Kan YM,Wang PL.Oxidation of ZrB2powder in thetemperature range of650–800°C.J Alloys Compd2009;471:502–6.[20]Han JC,Hu P,Zhang XH,Meng SH.Oxidation behavior of zirconium diboride-silicon carbide at1800°C.Scr Mater2007;57:825–8.[21]Chen DJ,Li WJ,Zhang XH,Hu P,Han JC,et al.Microstructural feature and thermalshock behavior of hot-pressed ZrB2–SiC–ZrO2composite.Mater Chem Phys 2009;116:348–52.[22]Han JC,Hu P,Zhang XH,Meng SH,Han WB.Oxidation-resistant ZrB2–SiCcomposites at2200°pos Sci Technol2008;68:799-06.48S.Meng et al./Int.Journal of Refractory Metals and Hard Materials29(2011)44–48。