Compressive creep behavior of hot-pressed Si 3N 4 ceramics using alumina and a rare earth solid solution as additives

Compressive creep behavior of hot-pressed Si 3N 4 ceramics using alumina and a rare earth solid solution as additives
of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Compressive creep behavior of hot-pressed Si 3 N 4  ceramicsusing alumina and a rare earth solid solution as additives C. Santos  a,b,* , K. Strecker  a , M.J.R. Barboza  a , F. Piorino Neto  b ,O.M.M. Silva  b , C.R.M. Silva  b a FAENQUIL, DEMAR, Polo Urbo Industrial, Gleba AI-6, s/n, Lorena-SP, CEP 12600-000, Brazil  b CTA-IAE, Pc¸a. Marechal do Ar Eduardo Gomes, 50, S.J. Campos-SP, CEP 12228-904, Brazil  Received 13 September 2004; accepted 24 February 2005 Abstract In this work, Si 3 N 4  based ceramics were hot-pressed using Al 2 O 3  and a solid solution of Y 2 O 3  and rare earth oxides (RE 2 O 3 )produced at DEMAR, FAENQUIL at a cost of approximately 25% of pure Y 2 O 3 , as additives. Two compositions with 5 and20 wt.% additive content at a constant ratio of 60 mol.% Al 2 O 3  to 40 mol.% RE 2 O 3  were investigated. The highly dense Si 3 N 4 ceramics were tested for their creep behavior in compression between 1235 and 1300   C under stresses ranging from 100 to350 MPa. Scanning electron microscopy and X-ray diffractometry were used to identify the predominantly acting creep mechanism.Higher additive amounts resulted in larger grains of higher aspect ratios and in a decreased anisotropy in the uniaxial hot-pressedceramic materials.The compressive creep behavior of the materials depends on the composition and amount of the intergranular phase. Whilehigher amounts of additives resulted in higher creep rates,  _ e , and higher stress exponents, the activation energy  Q ss , has been inferiorfor samples with lower additive contents,  n . This observation is attributed to the higher SiO 2  content due to the higher amountof Si 3 N 4  in the starting powders, resulting in liquid phase formation at lower temperatures. Grain sliding has been identified tobe the predominant mechanism responsible for creep deformation of these ceramics.   2005 Elsevier Ltd. All rights reserved. Keywords:  Si 3 N 4 ; Creep behavior; Oxidation resistance; Microstructure 1. Introduction Silicon nitride (Si 3 N 4 ) is a ceramic compound with anelevated degree of covalent bonding and high bondingenergy, resulting in a material of high hardness, strengthand elastic modulus. On the other hand, because of the bonding characteristics, self-diffusion coefficients areextremely small and additives must be used to promotesintering in order to achieve highly densified materials.Ingeneral,oxidessuchasAl 2 O 3 ,Y 2 O 3 andrareearthoxi-desareusedasadditives[1,2].Theadditivesformaliquidphaseatelevatedtemperaturesandpromotedensificationby the liquid phase sintering processes: rearrangement,solution of   a -Si 3 N 4  grains and precipitation as  b -Si 3 N 4 grains and Oswald ripening [2].At the Department of Materials Science and Engi-neering of FAENQUIL a process was developed forthe production of a mixed yttrium-rare earth oxide,RE 2 O 3 , where RE stands for rare earth elements, start-ing from ‘‘Xenotime’’, an yttrium rich phosphatic ore(RE,Y)PO 4 . This process is based on the alkaline fusion 0263-4368/$ - see front matter    2005 Elsevier Ltd. All rights reserved.doi:10.1016/j.ijrmhm.2005.02.003 * Corresponding author. Tel.: +55 12 31599915; fax: +55 1231533006. E-mail address: (C. Santos).International Journal of Refractory Metals & Hard Materials 23 (2005) 183–  of the ore with NaOH, followed by aqueous and acidleaching, precipitation by oxalic acid and subsequentcalcination [3]. The final product is a solid solution of  yttrium and rare earth (Er +3 , Yb +3 , Dy +3 , Ho +3 ) oxides,called RE 2 O 3 , where the rare earth ions substitute Y +3 inthe structure.The use of RE 2 O 3  as an alternative sinter additive forSi 3 N 4  has been studied because this material is obtainedat a lower cost compared to pure Y 2 O 3  and becausethe pure rare earth oxides are also viable additives,forming refractory intergranular phases. The substitu-tion of Y 2 O 3  by RE 2 O 3  resulted in comparable resultsof densification, microstructure and mechanical proper-ties [4–7].Due to their characteristics previously mentioned, thetechnological applications of silicon nitride ceramics(Si 3 N 4 ) are numerous, and in this context, the studyand understanding of the mechanical properties of thesematerials at high temperatures is of primary interest. Ithas also been demonstrated that the secondary glassyphase, which appears during the sintering, has a stronginfluence on the creep behavior.Hot-pressed silicon nitride is generally fabricatedusing oxide additives to enhance the densification. How-ever, the remaining glassy phase softens at high temper-atures and results in a deleterious effect on the hightemperature strength. The creep deformation of materi-als containing an intergranular glassy phase is generallythought to occur in combination of viscous flow, solu-tion-precipitation and cavitation mechanisms [8–12].Creep deformation of Si 3 N 4  could not be directly attrib-uted to a single factor such as additive content and com-position, grain size, etc., since the grain size is alsodependent on the composition and amount of sinteringadditives. Thus, the creep behavior of silicon nitrideceramics depends on many factors, such as the amountand composition of the glassy phase, the size distribu-tions of the matrix Si 3 N 4  grains, the partial recrystalliza-tion of the grain boundary glassy phase and oxidationresistance [13,14].The compressive creep behavior of Y 2 O 3  –Al 2 O 3 doped Si 3 N 4  ceramics obtained by hot-pressing havebeen studied by various authors [9,13,12]. Testing hasbeen performed in the temperature range between 1250and 1400   C, under stresses ranging from 30 to250 MPa. The results in terms of the stress exponent( n ) and the activation energy ( Q ss ) of the creep mecha-nism were quite similar, ranging between 0.85 and 1.5for the  n -values, and between 440 and 650 kJ/mol forthe  Q -values, depending on composition and processingconditions.The objective of the present work has been to inves-tigate the compressive creep behavior of Si 3 N 4  ceramicsdoped with Al 2 O 3  and an alternative sintering additive,an yttrium-rare earth oxide solid solution, in substitu-tion of the commonly applied pure-Y 2 O 3 . 2. Experimental procedure  2.1. Sample processing  Commercial  a -Si 3 N 4  (H.C. Starck, LC-12), Al 2 O 3 (Baikalox, CR-6) and an yttrium-rare earth oxide mix-ture RE 2 O 3  (DEMAR-FAENQUIL), were used asstarting powders. It has been shown in previous worksthat this oxide mixture is a solid solution of Y 2 O 3 (44 wt.%) with Yb 2 O 3  (17 wt.%), Er 2 O 3  (14 wt.%) andDY 2 O 3  (10 wt.%) as its major constituents, besidesminor amounts of Ho 2 O 3 , Tm 2 O 3 , Tb 2 O 3 , Lu 2 O 3 ,Gd 2 O 3  and Sm 2 O 3 . It can be used as an effective andcheap substitute for pure Y 2 O 3  as sinter additive forSi 3 N 4  ceramics, resulting in elevated relative densityand with comparable mechanical properties at roomtemperature [4,7].Two compositions were studied, containing 5 and20 vol.% of additives, with a constant molar ratio of 60% Al 2 O 3  to 40% RE 2 O 3 . Sample compositions inwt.% are listed in Table 1. The SiO 2  content has beencalculated using the manufacturer  s data sheet for theoxygen impurities of the Si 3 N 4  powder.Powder batches were produced by ball milling for 2 husing ethanol as milling media. After mixing, the pow-der batches were dried first in a rotary evaporator andsubsequently in an oven at 120   C for 12 h. Prior tohot-pressing the powder mixtures were sieved througha 60 mesh screen.Hot-pressing was done at 1750   C for 30 min under apressure of 20 MPa with a heating rate of 15   C/min innitrogen atmosphere, obtaining sintered specimen of approximately 25 mm diameter and 7 mm height. Fromthese discs, samples of 6  ·  3  ·  3 mm 3 were cut and usedfor creep testing. The densities of the sintered specimenswere determined by the immersion method in distilledwater, using Archimedes   principle and related to thetheoretical density, calculated by the rule of mixtures.Phase analysis was done by X-ray diffraction (XRD),both before and after the creep tests, comparing the dif-fraction patterns with the JCPDS files. X-ray diffractionanalysis was conducted on planes parallel and perpen-dicular to the hot pressing direction in a bulk specimen.From the X-ray diffraction patterns, the ratio of the(101)/(210) peak area of   b -Si 3 N 4  was used as an indi-cation of a preferential orientation of the elongated b -grains [13,16]. Table 1Composition of powder batchesDenomination Composition (wt.%)Si 3 N 4  Al 2 O 3  RE 2 O 3  SiO 2 SNCAL 5 89.58 2.85 5.15 2.42SNCAL 20 68.90 10.44 18.80 1.86184  C. Santos et al. / International Journal of Refractory Metals & Hard Materials 23 (2005) 183–192  For microstructural analysis, the specimens wereground, polished and chemically etched by a 1:1 mixtureof NaOH and KOH at 500   C for 3 min. The grain sizeand the aspect ratio were determined by the statisticalapproximation proposed by Wo¨tting et al. [19], who considered that 10% of the grains are exactly parallelto the plane analyzed and, consequently, indicate the‘‘real’’ aspect ratio of the grains. Approximately 1000 b -Si 3 N 4  grains have been measured, choosing the 10%of the apparently largest grains for determination of the average width, length and aspect ratio of the grains.The orientation of the grains was evaluated by determin-ing the angle  h  between the  c -axis of the  b -Si 3 N 4  grainsand the  x -axis of the planes normal and parallel to thehot-pressing direction [17], see Fig. 1.  2.2. Experimental apparatus used in creep testing  Compressive creep tests were carried in air using adead-weight-creep-rupture machine, at temperaturesranging between 1235 and 1300   C and under nominalstresses ranging between 100 and 350 MPa. The com-pressive load was transmitted to the sample by highcreep resistance SiC bars, and the length variation of the sample was measured using a linear variable differen-tial transformer (LVDT) with a precision of 1  l m. Thetemperature was controlled to ±2   C using a Pt/Pt– 10%Rh thermocouple.Samples of approximate 6  ·  3  ·  3 mm 3 were used inthe creep tests. The compressive creep behavior of thesamples has been investigated in the same axis as thehot-pressing direction. Due to the alignment of the elon-gated  b -Si 3 N 4  grains in planes perpendicular to the hot-pressing direction, an increasing creep resistance isexpected for this orientation, when compared to speci-men orientated perpendicular to the hot-pressing axis,because grain boundary sliding by viscous flow is moredifficult.Creep curves were obtained and the results analyzedin terms of the power law using the Norton equation, _ e ss  ¼  A r n exp   Q ss  RT     ð 1 Þ where  _ e ss  is the true steady state strain rate,  A  constant, r  the applied stress,  T   the absolute temperature,  R  thegas constant,  n  the stress exponent and  Q ss  the apparentactivation energy. The results were performed under twoconditions: at constant temperatures and different stress,in order to evaluate the stress exponent  n , and underconstant load and different temperatures, to evaluatethe apparent activation energy  Q ss . Fig. 2 shows theschematic apparatus used in the creep tests. 3. Results and discussion After hot-pressing, both compositions studied exhib-ited high relative densities of 98.5 ± 0.1% for SNCAL 5and 99.1 ± 0.3% for SNCAL 20, respectively. Phaseanalysis revealed only the presence of   b -Si 3 N 4  as crystal-line phase, indicating a complete transformation of   a -into  b -Si 3 N 4  and the formation of a secondary glassyphase by the additives. 3.1. Creep behavior—general considerations Previous works identified several creep mechanismsacting simultaneously in Si 3 N 4  ceramics. In these studiesthe predominant creep mechanism, have been identifiedto be grain sliding by viscous flux, solution-precipitationof the Si 3 N 4  grain and cavitation [12,18]. The simulta-neous occurrence of several creep mechanisms difficultthe understanding and precise description of the creepbehavior. Besides temperature and stress, the predomi-nant mechanism is determined by the type of testing,in tension, bending or compression. PHot pressingdirection x X P θ P Fig. 1. Schematic figure of a hot-pressed sample with preferential grain orientation in the plane parallel to the hot pressing direction (P).  h P represents the angle of grain orientation in the P plane. C. Santos et al. / International Journal of Refractory Metals & Hard Materials 23 (2005) 183–192  185  In general, creep deformation of Si 3 N 4  occurs athigher deformation rates in tensile testing by cavitation,when compared to testing in compression where grainsliding and/or solution-precipitation are the dominantmechanism. In the presence of some residual  a -Si 3 N 4 after sintering, temperatures close to 1400   C wherethe  a - to  b -Si 3 N 4  phase transformation starts, solu-tion-precipitation creep mechanism is favored.Besides that, a series of other factors, such as grainsize and morphology [15], preferential orientation of grains in hot-pressed materials, as well as amount, chem-icalcompositionandcrystallinityofintergranularphasesdetermine the creep behavior of these ceramics [13].Table 2 summarizes the results of the steady statecreep rates of samples SNCAL 5 and SNCAL 20, at dif-ferent temperatures and compressive stresses applied.As can be noted, creep rates increase with increasingstress and temperature, for both sample compositionsstudied. Creep curves for the highest stress level,300 MPa, and varying temperatures, as well as creepcurves at the highest temperature and varying stressesare shown in Figs. 3 and 4 for samples SNCAL 5 andSNCAL 20, respectively.From these curves and from the data presented inTable 2 it can be seen, that the creep rates are substan-tially higher for samples SNCAL 20, containing a muchhigher amount of additives. Furthermore, it can be seenfrom Figs. 3 and 4, that under the applied conditionsonly samples SNCAL 20 reached the third stage of creepdeformation, eventually leading to failure.Phase analysis of the sintered samples identified only b -Si 3 N 4  as crystalline phase, indicating that the additives Fig. 2. Schematic compressive creep tests apparatus.Table 2Creep results of the hot pressed samplesComposition Tests conditions Steady state creeprate ( e ss ) (h  1 )Temperature (  C) Stress (MPa)SNCAL 5 1250 200 2.50  ·  10  4 250 2.77  ·  10  4 300 3.39  ·  10  4 1275 200 5.64  ·  10  4 250 6.98  ·  10  4 300 8.02  ·  10  4 1300 100 6.04  ·  10  4 200 7.94  ·  10  4 250 9.58  ·  10  4 300 1.19  ·  10  3 350 2.18  ·  10  3 SNCAL 20 1235 200 2.45  ·  10  4 250 4.34  ·  10  4 300 6.98  ·  10  4 1275 200 1,31  ·  10  3 250 1.90  ·  10  3 300 3.42  ·  10  3 1300 100 2.15  ·  10  3 200 7.16  ·  10  3 250 9.43  ·  10  3 300 1.47  ·  10  2 186  C. Santos et al. / International Journal of Refractory Metals & Hard Materials 23 (2005) 183–192  employed formed a liquid phase during sintering andsolidified in an amorphous intergranular phase duringcooling. The type and amount of this secondary phaseis responsible to a high degree for the observed creepbehaviors of these materials.Samples sintered with 20 vol.% of additives, SNCAL20, exhibited an almost 10 times higher creep rate, whencompared to the samples containing only 5 vol.% addi-tives, SNCAL 5; the respective creep rates measuredbeing 1.47  ·  10  2 h  1 for SNCAL 20 and 1.19  ·  10  3 h  1 for SNCAL 5. 3.2. Microstructural aspects X-ray diffraction patterns of samples SNCAL 5 andSNCAL 20, before and after creep testing at 1300   Cunder a stress of 300 MPa during 60 and 15 h respec-tively, are shown in Fig. 5. The analysis was done ongrinded and polished surfaces in planes parallel to thehot-pressing axis, because of the alignment of the elon-gated Si 3 N 4  grains in this plane.As can be noted, samples SNCAL 5, did not exhibitsignificant variations in the diffraction patterns beforeand after creep testing, while a partial crystallizationof the intergranular phase forming YAG (Y 3 Al 5 O 12 )phase is evident for the samples SNCAL 20. In a pre-vious work the principle mechanism responsible forthe formation of texture, have been identified beinggrain rotation and preferential grain growth due to thepressure gradient [13].The alignment of grains during creep testing has beenstudied in greater detail by the variation of the (101) to(210) peak intensity ratio of   b -Si 3 N 4  in the planesparallel P and normal N to the hot-pressing direction.In Fig. 6 X-ray diffraction patterns of sample SNCAL5 for planes normal, N, and perpendicular, P, to thehot-pressing direction are shown prior and after creeptesting at 1300   C for 60 h under 300 MPa.While for the N plane no variations of the (101) to(210) peak intensity ratio of about 0.40 has been ob-served, this ratio increases from 1.40 to 1.75 for P plane,indicating that first anisotropy is much more accentu-ated in the P-plane and secondly that further grainsalignment occurs in this plane. Equivalent findings wereobtained for samples SNCAL 20, resulting in a constantpeak intensity ratio of about 0.45 for N plane and anincreasing ratio from 1.45 to 1.74, for the P plane. InFig. 7 the development of the peak intensities ratios 01020304050600. (a)    S   t  r  a   i  n   (  m  m   /  m  m   )   S   t  r  a   i  n   (  m  m   /  m  m   ) Time (h) 0102030405060 Time (h)SNCAL 5 1300 ° C - 200 MPa1275 ° C - 200 MPa1250 ° C- 200 MPa (b)SNCAL 5 1275 ° C - 200 MPa1275 ° C - 250 MPa1275 ° C - 300 MPa Fig. 3. Creep deformation of sample SNCAL 5: (a) submitted to 200 MPa, at different temperatures, and (b) at 1275   C, under different stresses. 01020304050600. (a)    S   t  r  a   i  n   (  m  m   /  m  m   )   S   t  r  a   i  n   (  m  m   /  m  m   ) Time (h) 0102030405060 Time (h)SNCAL 20 1235 ° C - 250 MPa1275 ° C - 250 MPa1300 ° C - 250 MPa (b)SNCAL 20 1300 ° C - 100 MPa1300 ° C - 200 MPa1300 ° C - 300 MPa Fig. 4. Creep deformation of sample SNCAL 20: (a) submitted to 250 MPa, at different temperatures, and (b) at 1300   C, under different stresses. C. Santos et al. / International Journal of Refractory Metals & Hard Materials 23 (2005) 183–192  187
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks