J.Am. Ceran.So,9图81817-1822(2009) DOl:10.l11551-2916.2009.03114 C)2009 The American Ceramic Society urna Fracture Toughness Analysis of Advanced Ceramic Composite for Hip Prosthesis Giuseppe pezzotti, .+. Kiyotaka Yamada, Alessandro Alan Porporati, Meinhard Kuntz, and Kengo yamamoto CEramic Physics Laboratory& Research Institute for Nanoscience, Kyoto Institute of Technology, Kyoto 606-8585, Japan Department of Orthopaedics, Orthopaedic Research Center, Loma Linda University, Loma Linda, California 92354 Piezotech Japan Ltd K yoto 606-8126, Japan Medical Products Division, CeramTec AG, Stuttgart, Plochingen D-73207, Germany Department of Orthopaedic Surgery, Tokyo Medical University, Tokyo 160-0023 Japan The fracture behavior of a zirconia-toughened alumina-matrix aration processes occur in heavy-weighted patients, as well as in composite(added with small amounts of mixed oxides) for ce- Asian patients, whose squatting and setting attitudes require a ramic hip joint prostheses has been evaluated with emphasis prosthesis design with high joint deflection. Furthermore, such placed on the effect of environmental surface degradation in a superior fracture behavior should be retained for long life- moist environment. Accelerated aging tests were performed times, as expected in young patients, by whom ceramic-on-ce- to 300 h in an autoclave operating at 121 C(under 0.1 MPa ramic joints shall be selected for obtaining an enhanced activity Dr envir which represents a quite severe In this scenario, much expectation is placed on the development environmental testing condition Besides conventional fracture of new ceramic materials, especially ceramic/cer mechanics characterizations, including different types of frac- ites, whose fracture behavior and reliability could be signi ture toughness test, microscopic insight into the effect of envi- ntly improved above those of monolithic ceramics ronmental surface degradation on toughness could be obtained Polycrystalline alumina possesses a very low coefficient of friction, negligible wear rate and extreme chemical inertness in The main outcomes of this study were as follows: after 10-h physiological environment, and has emerged as a very suitable autoclaving (according to Iso standard recommendation)n bearing material for ceramic orthopedic joints. However, con- significant change of monoclinic volume fraction and fracture cerns arise from its brittleness and, thus vulnerability to fracture toughness could be detected; (ii)after very long exposure time due to unexpected load situation. In order to counteract brit (300 h)the monoclinic phase content increased and the surface tleness, a dispersion of zirconia particles can be introduced in the fracture toughness decreased by approximately 30%, although alumina matrix to obtain a composite material, whose fracture it was still above the tour level of pure alumina; and (iithe resistance is greatly enhanced by the effect of a phase-transfor bulk toughness was unaffected by autoclave exposure, indepen- mation process occurring in the zirconia dispersoid dent of exposure time elongation It is the main object of this paper to evaluate the effectiveness of such a toughening mechanism in an alumina zirconia com- posite belonging to the next generation of ceramic materials for hip joints, and to quantitatively evaluate to which extent tough Moson et adern icuipatintcyrastdes s hensiat en towa head ening can be retained after exposing the material to long-term aging in moist environment as its sliding counterpart. During the last three decades, oxide- based ceramics have been increasingly introduced in orthopedic urgery in replacement of metallic cups and heads. This trend II. Experimental Procedure arises from clear advantages related to the superior biocompat In a forthcoming paper, 3 we have given the details of the an- ibility and sliding behavior of ceramic materials-3 and has led to alyzed composite material(BIOLoX"delta, Ceram Tec, Ploc- the development of ceramic-on-ceramic hip prostheses. 4 The main advantage in adopting a fully ceramic hip prosthesis is the hingen, Germany), consisting of 80 vol% Al,O3, 17 vol% ZrO, omplete elimination of contamination by polymeric debris and 3 vol% strontium aluminate. Y and Cr are solved in the leading to osteolysis and eventually to revision surgery)as irconia and alumina /aluminate matrix, respectively. Its inert wear subproducts from polyethylene-made acetabular cups ess in a moist environment has also been precisely character ized. Briefly here, aging periods up to several thousands of from the joint articulation and the ceramic femoral head directly hours in the range of temperature 90% C were conducted impinges on an equally hard acetabular cup, superior fracture a moist environment. According to these data. the activation istance/reliability becomes a necessary requirement for the energy for environmental transformation in the investigated lected ceramic. This is particularly true when joint microsep- erial was obtained as 108 kJ/mol. By considering such an activation energy value for the composite, I h at 134.C shall correspond to 3.9-year exposure in vivo. Recent studies3, have T. Troczynskicontributing editor shown that tribolayer formation and composition during sliding of alumina bearings in wet environment strongly depend on the pH of the liquid in which sliding P ccurs. In this study, the moist manuscript No 25815 Received January 30, 2009: approved March 21 2009 could be made to alter such conditions 1817
Fracture Toughness Analysis of Advanced Ceramic Composite for Hip Prosthesis Giuseppe Pezzotti,w,z,y Kiyotaka Yamada,z Alessandro Alan Porporati,z Meinhard Kuntz,J and Kengo Yamamoto z Ceramic Physics Laboratory & Research Institute for Nanoscience, Kyoto Institute of Technology, Kyoto 606-8585, Japan y Department of Orthopaedics, Orthopaedic Research Center, Loma Linda University, Loma Linda, California 92354 z Piezotech Japan Ltd., Kyoto 606-8126, Japan J Medical Products Division, CeramTec AG, Stuttgart, Plochingen D-73207, Germany Department of Orthopaedic Surgery, Tokyo Medical University, Tokyo 160-0023, Japan The fracture behavior of a zirconia-toughened alumina-matrix composite (added with small amounts of mixed oxides) for ceramic hip joint prostheses has been evaluated with emphasis placed on the effect of environmental surface degradation in moist environment. Accelerated aging tests were performed up to 300 h in an autoclave operating at 1211C (under 0.1 MPa pressure) in vapor environment, which represents a quite severe environmental testing condition. Besides conventional fracture mechanics characterizations, including different types of fracture toughness test, microscopic insight into the effect of environmental surface degradation on toughness could be obtained according to Raman and fluorescence microprobe spectroscopy. The main outcomes of this study were as follows: (i) after 10-h autoclaving (according to ISO standard recommendation) no significant change of monoclinic volume fraction and fracture toughness could be detected; (ii) after very long exposure time (300 h) the monoclinic phase content increased and the surface fracture toughness decreased by approximately 30%, although it was still above the toughness level of pure alumina; and (iii) the bulk toughness was unaffected by autoclave exposure, independent of exposure time elongation. I. Introduction MOST of modern hip joint prostheses consist of two components: an articulating cup and a spherical femoral head as its sliding counterpart. During the last three decades, oxidebased ceramics have been increasingly introduced in orthopedic surgery in replacement of metallic cups and heads. This trend arises from clear advantages related to the superior biocompatibility and sliding behavior of ceramic materials1–3 and has led to the development of ceramic-on-ceramic hip prostheses.4 The main advantage in adopting a fully ceramic hip prosthesis is the complete elimination of contamination by polymeric debris (leading to osteolysis and eventually to revision surgery) as wear subproducts from polyethylene-made acetabular cups. However, once the softer polyethylene counterpart is eliminated from the joint articulation and the ceramic femoral head directly impinges on an equally hard acetabular cup, superior fracture resistance/reliability becomes a necessary requirement for the selected ceramic. This is particularly true when joint microseparation processes occur in heavy-weighted patients,5 as well as in Asian patients, whose squatting and setting attitudes require a prosthesis design with high joint deflection.6 Furthermore, such a superior fracture behavior should be retained for long lifetimes, as expected in young patients, by whom ceramic-on-ceramic joints shall be selected for obtaining an enhanced activity. In this scenario, much expectation is placed on the development of new ceramic materials, especially ceramic/ceramic composites,7–10 whose fracture behavior and reliability could be signifi- cantly improved above those of monolithic ceramics. Polycrystalline alumina possesses a very low coefficient of friction, negligible wear rate and extreme chemical inertness in physiological environment, and has emerged as a very suitable bearing material for ceramic orthopedic joints. However, concerns arise from its brittleness and, thus, vulnerability to fracture due to unexpected load situation. In order to counteract brittleness, a dispersion of zirconia particles can be introduced in the alumina matrix to obtain a composite material, whose fracture resistance is greatly enhanced by the effect of a phase-transformation process occurring in the zirconia dispersoids.11,12 It is the main object of this paper to evaluate the effectiveness of such a toughening mechanism in an alumina/zirconia composite belonging to the next generation of ceramic materials for hip joints, and to quantitatively evaluate to which extent toughening can be retained after exposing the material to long-term aging in moist environment. II. Experimental Procedure In a forthcoming paper,13 we have given the details of the analyzed composite material (BIOLOXs delta, CeramTec, Plochingen, Germany), consisting of 80 vol% Al2O3, 17 vol% ZrO2, and 3 vol% strontium aluminate. Y and Cr are solved in the zirconia and alumina/aluminate matrix, respectively. Its inertness in a moist environment has also been precisely characterized.13 Briefly here, aging periods up to several thousands of hours in the range of temperature 901–1421C were conducted in a moist environment. According to these data, the activation energy for environmental transformation in the investigated material was obtained as 108 kJ/mol. By considering such an activation energy value for the composite, 1 h at 1341C shall correspond to 3.9-year exposure in vivo. Recent studies13,14 have shown that tribolayer formation and composition during sliding of alumina bearings in wet environment strongly depend on the pH of the liquid in which sliding occurs. In this study, the moist environment was in a near-neutral pH condition and no attempt could be made to alter such conditions. T. Troczynski—contributing editor w Author to whom correspondence should be addressed. e-mail: pezzotti@kit.ac.jp Manuscript No. 25815. Received January 30, 2009; approved March 21, 2009. Journal J. Am. Ceram. Soc., 92 [8] 1817–1822 (2009) DOI: 10.1111/j.1551-2916.2009.03114.x r 2009 The American Ceramic Society 1817
1818 Journal of the American Ceramic Society-Pezzonti et al. Vol. 92. No 8 enter to obtain a single-edge V-notch beam(SEVNB) config- uration for fracture mechanics assessments. In order to achieve stable fracture initiation in bending geometry, we used the crack stabilizer designed by Nojima and Nakai, 9 for a three- Confocal probe oint bending geometry. The span of the flexure jig was 16 mm and the cross-head speed was 0.005 mn the load-displacement relation was directly measured with eral semiconductor gauges, which were placed both on the load ing bar and on the tensile surface of the specimen. In low cross- head speed fracture experiments, the fracture toughness value relative to crack propagation from the notch root was calculated according to the equation given by Rawley. For comparison bending bar samples containing a notch with the same geometry Fig 1. Schematic describing a nondestructive method for measurin were also unstably fractured at the relatively fast speed of subsurface transformation profiles by confocal Raman spe 0.5 mm /min and the Kic value calculated from the recorded fracture load. In this latter case, the upper and lower spans of the four-point bending jig were 20 and 36 mm, respectively. In rinciples and experimental procedures for Raman and all bending tests, the notch configuration showed a root radius fluorescence spectroscopic evaluations have also been described 5sps8 um All fracture data shown here represent the average previously and, for brevity's sake, will not be repeated here. value of 7 tested samples per each testing conditions. Both types An additional Raman spectroscopy procedure was adopted for of fracture test, by indentation and sevnb methods, were per nondestructively and quantitatively probing the amount of formed on the as-received material and on the same material phase transformation(after long-term exposure in vapor envi- fter short-as well as long-term aging in autoclave. In the cal- ronment)along the material subsurface. This Raman spectro- culation of toughness values by various methods, a common scopic procedure involved a systematic translation of the optical value of Youngs modulus and Poissons ratio of the composite confocal probe along the material subsurface to determine the material were taken as E= 355 GPa and v=0. 24, respectively in-depth profile of transformed phase fraction. A schematic of s measured by strain-gauges in bending. A comparison was this confocal procedure in the presence of a surface phase-trans made between the toughness behaviors of the composite and of formation layer is given in Fig. 1. The full details of the confocal monolithic alumina obtained from the same maker (Blo- probe configuration and of the adopted probe deconvolution procedure have been given elsewhere. 6. 17 Raman and fuorescence spectroscopy procedures were also Indentation prints with a pyramidal indenter were introduced plied for quantitatively probing the amount of on the surface of the composite material before and after ransformation and the related microscopic stres were developed around the path of cracks arreste crit propagated from the corners of the print, were observed by a ical Kic value field-emission-gun scanning electron microscope(FEG-SEM Computational procedures involving the deconvolution of JSM-6500F, JEOL Ltd, Tokyo, Japan)and quantitatively an- the confocal optical probe were carried out with the help of alyzed with respect to their crack-opening displacement(COD commercially available software and crack-tip stress field. Samples were coated with a thin (5-nm-thick) layer of gold before observation. Near-tip COD assessments were used for assessing the surface fracture tough- ness of the composite material before and after exposure in au IlL. Experimental Results toclave. Six to 10 COd profiles were analyzed for each aging (1) Surface Fracture Toughness Versus Aging in Moist condition. For comparison, toughness measurements were also made from indentation crack length, according to a standard The impact of environmentally driven phase transformation on method. In both types of toughness measurement from inden- the material (surface) toughness was evaluated by recording tation crack, we used indentation prints obtained on finely pol- Cod profiles from indentation cracks introduced in the material ished portions of surface (obtained from the manufactured before and after environmental aging test. Indentation cracks femoral heads) with a load of 98 N. The salient parameters can be considered to be near-surface equilibrium cracks, which elated to the measurement of toughness by indentation crack thus are stably arrested at about their threshold stress intensity length method are given in Table I value for crack propagation, KIc. Systematic COD measure- The bulk toughness of the material was assessed with frac- ments of such cracks may give direct access to the inherent turing bending bars, which contained a controlled flaw Speci- toughness of the near-surface portion of the material. In mens for fracture mechanics testing were rectangular bending addition, such a toughness value may represent the crack-tip bars 3 mm x 4 mm x 20 mm in dimension. A sharp V-shaped toughness of the material. The Kic value can be retrieved from notch with a relative length, a/Wa0.4, was introduced at the the slope of a plot of COD versus 8b F(a,x)/TE.Where Table L. Toughness Data by Different Testing Methods for the Composite in the as-Received State and After Exposure in autoclave at 121 C for Increasing periods of time KC(MPa. 2) IFCL (0.005mm/min) As received 6.5+0.6 3.2+0.2 3.0+0.2 Aged for 10 h 6.5+0.7 29+0.2 Aged for 300 h 4.7+0.5 2.2+0.2 2.8+0.2 5.6±04 Monolithic alumina 3.4+0.8 2.2+0.2 2.2+03 0.3 COD, crack-opening displacement: SEVNB single-edge V-notch beam
Principles and experimental procedures for Raman and fluorescence spectroscopic evaluations have also been described previously15 and, for brevity’s sake, will not be repeated here. An additional Raman spectroscopy procedure was adopted for nondestructively and quantitatively probing the amount of phase transformation (after long-term exposure in vapor environment) along the material subsurface. This Raman spectroscopic procedure involved a systematic translation of the optical confocal probe along the material subsurface to determine the in-depth profile of transformed phase fraction. A schematic of this confocal procedure in the presence of a surface phase-transformation layer is given in Fig. 1. The full details of the confocal probe configuration and of the adopted probe deconvolution procedure have been given elsewhere.16,17 Indentation prints with a pyramidal indenter were introduced on the surface of the composite material before and after exposure to moist environment. Indentation cracks, which propagated from the corners of the print, were observed by a field-emission-gun scanning electron microscope (FEG-SEM; JSM-6500F, JEOL Ltd., Tokyo, Japan) and quantitatively analyzed with respect to their crack-opening displacement (COD) and crack-tip stress field. Samples were coated with a thin (5-nm-thick) layer of gold before observation. Near-tip COD assessments were used for assessing the surface fracture toughness of the composite material before and after exposure in autoclave. Six to 10 COD profiles were analyzed for each aging condition. For comparison, toughness measurements were also made from indentation crack length, according to a standard method.18 In both types of toughness measurement from indentation crack, we used indentation prints obtained on finely polished portions of surface (obtained from the manufactured femoral heads) with a load of 98 N. The salient parameters related to the measurement of toughness by indentation cracklength method are given in Table I. The bulk toughness of the material was assessed with fracturing bending bars, which contained a controlled flaw. Specimens for fracture mechanics testing were rectangular bending bars 3 mm 4 mm 20 mm in dimension. A sharp V-shaped notch with a relative length, a/W0.4, was introduced at their center to obtain a single-edge V-notch beam (SEVNB) configuration for fracture mechanics assessments. In order to achieve stable fracture initiation in bending geometry, we used the crack stabilizer designed by Nojima and Nakai,19 for a threepoint bending geometry. The span of the flexure jig was 16 mm and the cross-head speed was 0.005 mm/min. During loading, the load–displacement relation was directly measured with several semiconductor gauges, which were placed both on the loading bar and on the tensile surface of the specimen. In low crosshead speed fracture experiments, the fracture toughness value relative to crack propagation from the notch root was calculated according to the equation given by Srawley.20 For comparison, bending bar samples containing a notch with the same geometry were also unstably fractured at the relatively fast speed of 0.5 mm/min and the KIC value calculated from the recorded fracture load. In this latter case, the upper and lower spans of the four-point bending jig were 20 and 36 mm, respectively. In all bending tests, the notch configuration showed a root radius 5rrr8 mm. All fracture data shown here represent the average value of 7 tested samples per each testing conditions. Both types of fracture test, by indentation and SEVNB methods, were performed on the as-received material and on the same material after short- as well as long-term aging in autoclave. In the calculation of toughness values by various methods, a common value of Young’s modulus and Poisson’s ratio of the composite material were taken as E 5 355 GPa and n 5 0.24, respectively, as measured by strain-gauges in bending. A comparison was made between the toughness behaviors of the composite and of monolithic alumina obtained from the same maker (BIOLOXs forte, CeramTec). Raman and fluorescence spectroscopy procedures were also applied for quantitatively probing the amount of zirconia phase transformation and the related microscopic stress fields, which were developed around the path of cracks arrested at their critical KIC value. Computational procedures involving the deconvolution of the confocal optical probe were carried out with the help of commercially available software.21 III. Experimental Results (1) Surface Fracture Toughness Versus Aging in Moist Atmosphere The impact of environmentally driven phase transformation on the material (surface) toughness was evaluated by recording COD profiles from indentation cracks introduced in the material before and after environmental aging test. Indentation cracks can be considered to be near-surface equilibrium cracks,22 which thus are stably arrested at about their threshold stress intensity value for crack propagation, KIC. Systematic COD measurements of such cracks may give direct access to the inherent toughness of the near-surface portion of the material. In addition, such a toughness value may represent the crack-tip toughness of the material. The KIC value can be retrieved from the slope of a plot of COD versus 8b1/2F(a, x)/pE0 . Where Fig. 1. Schematic describing a nondestructive method for measuring subsurface transformation profiles by confocal Raman spectroscopy. Table I. Toughness Data by Different Testing Methods for the Composite in the as-Received State and After Exposure in Autoclave at 1211C for Increasing Periods of Time Material KIC (MPa m1/2) IFCL COD c/a E/H SEVNB (0.005 mm/min) (0.5 mm/min) As received 6.570.6 3.270.2 3.070.2 5.770.6 3.4 0.2 Aged for 10 h 6.570.7 3.170.1 2.970.2 3.4 0.2 Aged for 300 h 4.770.5 2.270.2 2.870.2 5.670.4 3.8 0.2 Monolithic alumina 3.470.8 2.270.2 2.2703 4.8 0.3 COD, crack-opening displacement; SEVNB, single-edge V-notch beam. 1818 Journal of the American Ceramic Society—Pezzotti et al. Vol. 92, No. 8
August 2009 Advanced Ceramic Composite for Hip Prosthesis 1819 E=E/(1-v) is the plane strain elastic modulus; x, a, and b are an abscissa along the crack path( behind the crack tip and with 3.1 MPam origin at the tip). the distance between the crack tip and K-22 MP. mn the geometrical center of the indentation print, and the half diagonal of the indentation print, respectively. F(a, x)is a a Monolithic al weight function for the median surface crack configuration. as 且 Ritchie en by Fett et al. 2 and recently applied by Kruzic and 80.15 K"3.2 MPa Craclgure 2(a)shows FEG-SEM images of the main indentation crack path in a zone of relatively high COD values, while Fig 2(b) shows a microscopic branched area in the very neigh- Kx=22 MPam borhood of the crack tip. Both these figures refer to the material in the as-received state. Similar FEG-SEM pictures were used 0600 throughout the investigation for COD assessments in the as- F(a, x) received and autoclave-aged material. In Fig. 3, cod plots are the composite in the as-received state, after 10 and 300 h aging in cracks as collected in the as-received material and in the material au- moist atmosphere. From the best-fit sl f the plots, KI oclaved for 300 h at 121C. The slopes of the plots represent the crack values equal to 3. 2, 3. 1, and 2.2 MPa m/ were found for the tip toughness of the materials. as-received, 10-h, and the 300-h autoclaved material, respec- tively. The COd assessments show that no effect occurs on showed a surface toughness value similar to that of the com- toughness after 10-h aging, but also suggest that, ultimately, a posite after very long-term exposure in moist environment reduction in toughness by about one-third will occur after a very long-term exposure in autoclave. Note that the reduction in oughness observed in this experiment may also represent the actual contribution of phase transformation to toughening in (2) Bulk Toughness Versus Aging in Moist Atmosphere the as-received material. as discussed in a later section. Table I The results of fracture mechanics tests performed as a function nows a comparison between Kic values obtained from COD of stable crack extension, a, from a sharp notch root are shown measurements and those calculated according to the standard able I In such low cross-head speed fracture experi dentation method (namely, from indentation crack-length he critical stress intensity factor for crack initiation from the asurements). From a comparison between the two meth- otch root was KIc=3.0±0.24,2.9±0.20,and28±0.18 ds, it is clear that Kic values obtained from indentation MPa. 2. for the as-received sample and the samples aged in crack-length measurements were systematically larger than autoclave for 10 and 300 h, respectively. a detectably higher hose obtained from COD measurements. An explanation of fracture resistance was noticed when notched samples were frac this discrepancy will be given in Section IV(1). However, it is tured at high cross-head speed(cf. data listed in Table I); ho nportant to note here that, despite the observed discrepancy ever, again no significant difference in the toughness value could n the absolute values of toughness, a similar reduction in be found before and after autoclaving the samples. In other toughness upon autoclaving was found after both COd and in- ords, even long-term autoclaving does not appear to signifi- dentation crack-length measurement methods. In Table I,a antly affect the bulk toughness of the composite. In fracturing arison is also given with monolithic alumina, which notched bars at relatively fast cross-head speed, the observed Kic value was almost twice that measured at low cross-head speed a qualitative explanatio difference will be given the next section. Here, it is important to note that, as far as the bulk toughness value is concerned, the composite remains tougher than monolithic alumina even after very long-term ex- posures in moist environment (cf. Table D). In the present com- posite, a conspicuous toughening contribution is expected to irise from the residual stress fields associated with the tetrago- nal-to-monoclinic polymorphic transformation of zirconia - COD dispersoids in the neighborhood of the crack path. This phe- nomenon is likely to give a primary contribution to the tough- ening behavior In Section IV(2), we shall show a set of Raman also be attempted in Section IV(3) to provide a rationale for the bserved invariance of bulk toughness data upon aging in moist IV. Discussion (1) Rationale for the Discrepancy in Tou Measured by Di∥ erent In the examination of toughness data collected on the present 1 um composite material by various testing methods(as listed in Table D), large differences among absolute values obtained by different Fig. 2. Field-emission-gun scanning electron microscope images of the methods and or testing conditions can be noticed While this finding may stress the importance of a clear specification of the testing procedure adopted before comparing toughness data (COD)values;(b)the bridging zone in the neighborhood of the crack from different materials, it also calls for a clarification of the tip. Arrows here indicate COD displacements in(a) and bridging loca physical reasons behind the observed discrepancies in measuring tions in(b). toughness values. An attempt to interpret the reasons behind the
E0 5 E/(1n2 ) is the plane strain elastic modulus; x, a, and b are an abscissa along the crack path (behind the crack tip and with origin at the tip), the distance between the crack tip and the geometrical center of the indentation print, and the half diagonal of the indentation print, respectively. F(a, x) is a weight function for the median surface crack configuration, as first given by Fett et al.,23 and recently applied by Kruzic and Ritchie.24 Figure 2(a) shows FEG-SEM images of the main indentation crack path in a zone of relatively high COD values, while Fig. 2(b) shows a microscopic branched area in the very neighborhood of the crack tip. Both these figures refer to the material in the as-received state. Similar FEG-SEM pictures were used throughout the investigation for COD assessments in the asreceived and autoclave-aged material. In Fig. 3, COD plots are shown as collected on typical indentation cracks introduced in the composite in the as-received state, after 10 and 300 h aging in moist atmosphere. From the best-fit slopes of the plots, KIC values equal to 3.2, 3.1, and 2.2 MPa m1/2 were found for the as-received, 10-h, and the 300-h autoclaved material, respectively. The COD assessments show that no effect occurs on toughness after 10-h aging, but also suggest that, ultimately, a reduction in toughness by about one-third will occur after a very long-term exposure in autoclave. Note that the reduction in toughness observed in this experiment may also represent the actual contribution of phase transformation to toughening in the as-received material, as discussed in a later section. Table I shows a comparison between KIC values obtained from COD measurements and those calculated according to the standard indentation method (namely, from indentation crack-length measurements18). From a comparison between the two methods, it is clear that KIC values obtained from indentation crack-length measurements were systematically larger than those obtained from COD measurements. An explanation of this discrepancy will be given in Section IV(1). However, it is important to note here that, despite the observed discrepancy in the absolute values of toughness, a similar reduction in toughness upon autoclaving was found after both COD and indentation crack-length measurement methods. In Table I, a comparison is also given with monolithic alumina, which showed a surface toughness value similar to that of the composite after very long-term exposure in moist environment. (2) Bulk Toughness Versus Aging in Moist Atmosphere The results of fracture mechanics tests performed as a function of stable crack extension, a, from a sharp notch root are shown in Table I. In such low cross-head speed fracture experiments, the critical stress intensity factor for crack initiation from the notch root was KIC 5 3.070.24, 2.970.20, and 2.870.18 MPa m1/2, for the as-received sample and the samples aged in autoclave for 10 and 300 h, respectively. A detectably higher fracture resistance was noticed when notched samples were fractured at high cross-head speed (cf. data listed in Table I); however, again no significant difference in the toughness value could be found before and after autoclaving the samples. In other words, even long-term autoclaving does not appear to signifi- cantly affect the bulk toughness of the composite. In fracturing notched bars at relatively fast cross-head speed, the observed KIC value was almost twice that measured at low cross-head speed. A qualitative explanation for this difference will be given in the next section. Here, it is important to note that, as far as the bulk toughness value is concerned, the composite remains tougher than monolithic alumina even after very long-term exposures in moist environment (cf. Table I). In the present composite, a conspicuous toughening contribution is expected to arise from the residual stress fields associated with the tetragonal-to-monoclinic polymorphic transformation of zirconia dispersoids in the neighborhood of the crack path. This phenomenon is likely to give a primary contribution to the toughening behavior. In Section IV(2), we shall show a set of Raman spectroscopic results supporting the above argument. In addition, in-depth Raman assessments with a confocal probe will also be attempted in Section IV(3) to provide a rationale for the observed invariance of bulk toughness data upon aging in moist environment. IV. Discussion (1) Rationale for the Discrepancy in Toughness Values Measured by Different Methods In the examination of toughness data collected on the present composite material by various testing methods (as listed in Table I), large differences among absolute values obtained by different methods and/or testing conditions can be noticed. While this finding may stress the importance of a clear specification of the testing procedure adopted before comparing toughness data from different materials, it also calls for a clarification of the physical reasons behind the observed discrepancies in measuring toughness values. An attempt to interpret the reasons behind the Fig. 2. Field-emission-gun scanning electron microscope images of the profile of a crack propagated in the alumina/zirconia composite: (a) a main zone of the crack with relatively high crack-opening displacement (COD) values; (b) the bridging zone in the neighborhood of the crack tip. Arrows here indicate COD displacements in (a) and bridging locations in (b). Fig. 3. Crack-opening displacement (COD) plots for indentation cracks as collected in the as-received material and in the material autoclaved for 300 h at 1211C. The slopes of the plots represent the cracktip toughness of the materials. August 2009 Advanced Ceramic Composite for Hip Prosthesis 1819
1820 Journal of the American Ceramic Society--Pezzotti et al Vol. 92. No 8 observed differences in the toughness value is offered in the fol- loading speed Kic is measured, while Kio is likely the output of esting at slow loading speed thus involving the detrimental Experimental practice has shown us that the main difficulty effect of slow crack growth on the material crack resistance. in measuring toughness by indentation-related methods resides the case of the present material, the crack-shielding action is n the difficulty of precisely assessing the location of the crack believed to arise from the development of compressive residual tip in the present composite. As can be seen from the FEG-SEM stress field related to the phase-transformation zone in the neigh image shown in Fig. 2(b), the location of the tip on crack profile borhood of the crack path n the biolox delta material typically ends with a branched morphology (cf arrows in Fig. 2(b). The linear extension of this branched portion of crack profile may be as large as 10 um, (2)Spectroscopic Investigation of Su face Fracture which involves a significant difference in the calculated toug ness afi ness value, in the case of both COD and crack-length indenta Raman maps of monoclinic phase fraction are shown in Figs tion methods(referred to as IFCL in Table D). Excluding the 5(a)and(b), which were collected in the neighborhood of in- bridging zone from the considered crack length may lead to an dentation cracks propagated in the as-received material and in overestimation of the toughness value up to about 15%0-20% the material exposed for 300 h at 121C in autoclave, respec Toughness assessments by COD in the FEG-SEM22-24 are more tively. From a comparison between these two maps, it can be reliable, but involve the problem of using a weight function for a seen that a clear transformation zone is formed around the crack crack with a greatly altered geometry, as compared with a stan- path in the as-received ple, while in the aged sample the dard median crack configuration. This problem is visualized in surface crack runs in a fully transformed area, with almost no Fig 4, in which the same COd profile, as collected in the as possibility of newly creating an(additional) transformation zone received material, can be routinely best fitted by the same weight n its neighborhood. In Figs. 5(c)and(d), the corresponding function but selecting different crack-tip locations. The Kic val- maps of equilibrium stress(calculated as explained in the com- ues obtained for different crack-tip locations are also shown in panion paper) are shown for the sample before and after au- Fig. 4. As seen, if we incorporate the bridging zone in the CoD toclaving, respectively. Some features in these two additional profile, the routine leads to a poor fitting of the experiment maps may support the difference found in the measured KIc data in the neighborhood of the selected crack tip location; in values before and after autoclaving process. a broader zone of addition, a lower Kic value is found as compared with a fitting stress intensification is noticed at the tip of the equilibrium crack in which the crack tip location is selected by excluding the crack propagated in the autoclaved material as compared with that tip branched portion from the overall crack length. The tough- propagated in the as-received material(thus refile ness value calculated by this latter approach is closer to the value ence in the K-field to which the crack is subjected). Although the calculated by crack-length assessment (i.e, according to Japa absolute magnitude of the measured stresses can be affected nese Standards Association). An additional consideration (and underestimated) by the micrometer size of the laser probe about the relatively low value of fracture toughness detected it appears that crack-tip stress intensification presumabl by the Cod method can be stated according to the relatively spreads toward a larger area as compared with the as-received material, as a consequence of overlapping between the crack-tip COD experiments were made in the FE-SEM since the inden- stress field and the tensile residual stress field stored in the ma- terial surface after long-term aging in autoclave(cf. Pezzotti Ko (rather than a Kid) value is likely to be measured (i.e, Kol et al. ) By comparing the Raman spectroscopic maps in Fig.5 kc≈0.506)25 and the surface toughness data in Table I, it can be concluded The difference in toughness values by sEVNB methoc that, in the present composite, the main toughening effect (i.e, dopting stable and ble crack propagation can be mainas about 30% increase in toughness) may arise from the formation inferred to the significant difference in cross-head speed between of a phase transformation zone in the neighborhood of the crack the two testing procedures. It has been already reported in the path. Note that, in the same material investigated in this study,a literature2 that. in ally stabilized zirconia ceramics. th contribution to toughness arising from bridging effects by plate- toughness value in notched bars increased with increasing load- let-like grains has been also suggested. The present data instead ng speed (i.e, in the same interval of the present testing pro- suggest that the preponderant toughness effect arises from phase edures). In a previous 9 we have explained this transformation effects, which is also supported by the rather henomenon with the subcritical formation of a shielded (pop small fraction of platelet-like grains present in the material in) crack zone ahead of the notch root in semibrittle materials Again, the"rule"of kyo/ kic x0 5-0. 6 can be invoked to explain (a) 100% 100% Experimental data Excluding micro-cracking zone 0 0% Including micro-cracking zone Kie =4. 8 MPa.mn 200 MPa 200 MPa Kr-3.2 MPa.mm 戛 -200 MPa 200 MPa Distance from crack tip, x (um) Fig. 5. Raman maps of monoclinic phase fraction in the neighborhood Fig 4. Crack-opening displacement( CoD) data and two possible of the tip of indentation cracks, as propagated in the as-received material short-term, and long-term autoclaved alumina/zirconia composite, and (a) and in the material exposed for 300 h in autoclave at 121C(b).The ng criteria for the near-tip Cod profile as detected in the as-received corresponding maps of equilibrium stress are also shown in(c)and (d) for samples before and after autoclaving, respectively
observed differences in the toughness value is offered in the following. Experimental practice has shown us that the main difficulty in measuring toughness by indentation-related methods resides in the difficulty of precisely assessing the location of the crack tip in the present composite. As can be seen from the FEG-SEM image shown in Fig. 2(b), the location of the tip on crack profile in the BIOLOXs delta material typically ends with a branched morphology (cf. arrows in Fig. 2(b)). The linear extension of this branched portion of crack profile may be as large as 10 mm, which involves a significant difference in the calculated toughness value, in the case of both COD and crack-length indentation methods (referred to as IFCL in Table I). Excluding the bridging zone from the considered crack length may lead to an overestimation of the toughness value up to about 15%–20%. Toughness assessments by COD in the FEG-SEM22–24 are more reliable, but involve the problem of using a weight function for a crack with a greatly altered geometry, as compared with a standard median crack configuration. This problem is visualized in Fig. 4, in which the same COD profile, as collected in the asreceived material, can be routinely best fitted by the same weight function but selecting different crack-tip locations. The KIC values obtained for different crack-tip locations are also shown in Fig. 4. As seen, if we incorporate the bridging zone in the COD profile, the routine leads to a poor fitting of the experimental data in the neighborhood of the selected crack tip location; in addition, a lower KIC value is found as compared with a fitting in which the crack tip location is selected by excluding the cracktip branched portion from the overall crack length. The toughness value calculated by this latter approach is closer to the value calculated by crack-length assessment (i.e., according to Japanese Standards Association18). An additional consideration about the relatively low value of fracture toughness detected by the COD method can be stated according to the relatively long period of time (at least several hours) that elapsed until COD experiments were made in the FE-SEM since the indentation prints were made. Under such experimental conditions, a KI0 (rather than a KIC) value is likely to be measured (i.e., KI0/ KIC0.5–0.6).25–27 The difference in toughness values by SEVNB methods adopting stable and unstable crack propagation can be mainly inferred to the significant difference in cross-head speed between the two testing procedures. It has been already reported in the literature28 that, in partially stabilized zirconia ceramics, the toughness value in notched bars increased with increasing loading speed (i.e., in the same interval of the present testing procedures). In a previous paper,29 we have explained this phenomenon with the subcritical formation of a shielded (popin) crack zone ahead of the notch root in semibrittle materials. Again, the ‘‘rule’’ of KI0/KIC0.5–0.6 can be invoked to explain the different results25–27: in SEVNB testing conducted at fast loading speed KIC is measured, while KI0 is likely the output of testing at slow loading speed, thus involving the detrimental effect of slow crack growth on the material crack resistance. In the case of the present material, the crack-shielding action is believed to arise from the development of compressive residual stress field related to the phase-transformation zone in the neighborhood of the crack path. (2) Spectroscopic Investigation of Surface Fracture Toughness After Long-Term Aging Raman maps of monoclinic phase fraction are shown in Figs. 5(a) and (b), which were collected in the neighborhood of indentation cracks propagated in the as-received material and in the material exposed for 300 h at 1211C in autoclave, respectively. From a comparison between these two maps, it can be seen that a clear transformation zone is formed around the crack path in the as-received sample, while in the aged sample the surface crack runs in a fully transformed area, with almost no possibility of newly creating an (additional) transformation zone in its neighborhood. In Figs. 5(c) and (d), the corresponding maps of equilibrium stress (calculated as explained in the companion paper15) are shown for the sample before and after autoclaving, respectively. Some features in these two additional maps may support the difference found in the measured KIC values before and after autoclaving process. A broader zone of stress intensification is noticed at the tip of the equilibrium crack propagated in the autoclaved material as compared with that propagated in the as-received material (thus reflecting a difference in the K-field to which the crack is subjected). Although the absolute magnitude of the measured stresses can be affected (and underestimated) by the micrometer size of the laser probe, it appears that crack-tip stress intensification presumably spreads toward a larger area as compared with the as-received material, as a consequence of overlapping between the crack-tip stress field and the tensile residual stress field stored in the material surface after long-term aging in autoclave (cf. Pezzotti et al. 15). By comparing the Raman spectroscopic maps in Fig. 5 and the surface toughness data in Table I, it can be concluded that, in the present composite, the main toughening effect (i.e., about 30% increase in toughness) may arise from the formation of a phase transformation zone in the neighborhood of the crack path. Note that, in the same material investigated in this study, a contribution to toughness arising from bridging effects by platelet-like grains has been also suggested.7 The present data instead suggest that the preponderant toughness effect arises from phase transformation effects, which is also supported by the rather small fraction of platelet-like grains present in the material. Fig. 4. Crack-opening displacement (COD) data and two possible fitting criteria for the near-tip COD profile as detected in the as-received, short-term, and long-term autoclaved alumina/zirconia composite, and in monolithic alumina. Fig. 5. Raman maps of monoclinic phase fraction in the neighborhood of the tip of indentation cracks, as propagated in the as-received material (a) and in the material exposed for 300 h in autoclave at 1211C (b). The corresponding maps of equilibrium stress are also shown in (c) and (d) for samples before and after autoclaving, respectively. 1820 Journal of the American Ceramic Society—Pezzotti et al. Vol. 92, No. 8
August 2009 Advanced Ceramic Composite for Hip Prosthesis 182l (3) Spectroscopic Proof for Invariance of Bulk Toughness Table II. Optical Parameters for Different Component Upon Aging in Autoclave Phases In environmentally assisted phase-transformation prod polymorphic changes are maximized at the sample surfac eand ry quite steeply with tion along the in-depth axis. On the Alumina +tetragonal I mixture 1545 other hand. the raman is of a finite diameter and de Monoclinic zirconia 0.03 nd thus the collected at each location represents an average(and weighted) Raman emission over the entire prob olume. The main consequence of the finite size of the prol s that, even if the phase-fraction data are collected with a highly parameters from Table II in Eq (1), an iterative computational confocal iguration, they need to be treated by a routine was run using commercially available software to find mathematical deconvolution procedure to retrieve the actual in the parameters A, B, C, m, and n, which best fit the experi depth transformation profile. The profiles of monoclinic phas VMzo) values, as retrieved from the in-depth confocal line scan. fraction, VMzo), which were collected with an in-depth confocal The results of this mathematical deconvolution procedure are scanning along the subsurface of samples in the as-received con- plotted in Fig. 6. The deconvoluted monoclinic fraction profile. dition and autoclaved for 300 h, are shown in Fig. 6. In this plot. VMz), is also compared in Fig. 6 with the monoclinic fraction corresponds to the in-depth abscissa, which locates the probe profile typically found in the as-received material. This latter profile was almost unaltered after 10-h exposure in moist envi positive zo value refers here to a translation of the focal plan ronment. On the other hand, after exposures in autoclave a below the free surface of the sample. The plot, VM=VM(zo) long as 300 h, the thickness of material affected by an enhanced was translated into a plot, VM= VM(z), with z being the abscissa amount of monoclinic phase increased. This spectroscopic as- of the actual phase fraction distribution along the in-depth di- sessment may be invoked to explain the reason why the bulk rection. defined similarly to zo. In other words, the experimental toughness value was almost independent of autoclaving expo- profile of phase fraction was cleaned up from the averaging sure time. In fact, we notice that, whatever steep the variance in tion confined to the level of a few tens of micrometers even after very long-term exposure in autoclave. The thickness of the transfo M(20) exp(-2x(4+n2)M(2)P(4+n(2 tion layer is important because it might affect the residual exp(-2(4+n2)P(4+n(2,=0d stress fields developed on the bearing surface and, thus, the wear behavior of the material. It has been reported that a compres- sive residual stress on the bearing surface contributes to reduce P(=,z0 (2) both amounts of wear and friction in alumina materials.In +(2-20) Pezzotti et al.,we have quantitatively measured the residual stress in the alumina phase on the surface of the composite be- where the para 16. the subscripts in brackets locate whether ameters P, P, and Pe are the material /probe inter- e to moist environment. a trend w found showing an initial moderate decrease to reach a magni- the parameter refers to an alumina/ tetragonal zirconia mixture mixture tude of few tens of MPa in compression after about 20-h expo- r to the monoclinic zirconia phase(A+T and M, respectively) sure(transformation thickness of about 6 um) and then a Solving the integral Eq (1) for VM(z)involves the knowledge of gradual increase of the residual stress to reach zero magnitude the mathematical character of the in-depth profile function after about 45 h and toward the tensile side(with Based on previous Raman assessments from the sample cross 45 h). This may suggest that the alumina matrix in the pre tion to be structured as follows 6amics, we assume this func- sections in monolithic zirconia ce omposite could be more wear resistant than monolithic al VM()e Aexp(-Bx)+ In this study we hay acture where A, B, C, m, and n are constants, and Vo is the(initial) mechanics and confocal Raman microp volume fraction of monoclinic phase in the bulk of the as- characterize both bulk and surface toughness of an alumina/zir eceived material. After substituting for Eq. (3)and the probe conia composite used in advanced joint prostheses (e.g,BIO- LOX delta femoral heads). In a previous paper, we showed that transformation occurs in the material as a consequence of an environmentally assisted process. The autoclave testing p As-recerved cedure adopted in that study was quite severe and corresponded to exposures in human body of several lifetimes. The effect of an Aged for 10 h exposure of 10 h in autoclave at 134C was also examined 805050 hich, according to the activation energy for environmental re- sistance of the present composite, corresponds to a lifetime pectation of 39 years in vivo. Based on these results, the main 点A outcomes of the present fracture mechanics assessments can be ummarized as follows: (i) the crack-tip toughness level mea- ured in the as-received material is comprehensive of a tangible contribution by transformation toughening: (ii) even after the material is environmentally aged for periods of the order of 20 hundreds of hours and its surface toughness is reduced by about In-depth abscissa, z, or z (um) one-third the measured toughness of the composite measured Fig. 6. Profiles of monoclinic phase fraction, VAEo), as collected with y any method is at least the same as that of monolithic al- an in-depth confocal along the subsurface of the sample au- a:(ii) the observed decrease of fracture toughness by about best-fit convoluted curves and the cor. one-third is limited to the very surface of the material (i. e, to a responding deconvolut (calculated according to Eqs. (1H3)) layer of the order of a few tens of micrometers) and does not are also shown affect the bulk fracture toughness of the composite, as shown by
(3) Spectroscopic Proof for Invariance of Bulk Toughness Upon Aging in Autoclave In environmentally assisted phase-transformation processes, polymorphic changes are maximized at the sample surface and vary quite steeply with position along the in-depth axis. On the other hand, the Raman probe is of a finite diameter and depth, and thus the spectrum collected at each location represents an average (and weighted) Raman emission over the entire probe volume.14,15 The main consequence of the finite size of the probe is that, even if the phase-fraction data are collected with a highly confocal microprobe configuration, they need to be treated by a mathematical deconvolution procedure to retrieve the actual indepth transformation profile. The profiles of monoclinic phase fraction, VM(z0), which were collected with an in-depth confocal scanning along the subsurface of samples in the as-received condition and autoclaved for 300 h, are shown in Fig. 6. In this plot, z0 corresponds to the in-depth abscissa, which locates the probe focal plane, with origin at the material surface (cf. Fig. 1). A positive z0 value refers here to a translation of the focal plane below the free surface of the sample. The plot, VM 5 VM(z0), was translated into a plot, VM 5 VM(z), with z being the abscissa of the actual phase fraction distribution along the in-depth direction, defined similarly to z0. In other words, the experimental profile of phase fraction was cleaned up from the averaging effect of the probe upon solving the following integral equation16: VMðz0Þ ¼ R þ1 0 expð2að Þ AþT zÞVMðzÞPð Þ AþT ðz; z0Þdz R þ1 0 expð2að Þ AþT zÞPð Þ AþT ðz; z0Þdz (1) Pðz; z0Þ ¼ p2 c p2 1 þ p2 p2 c p2 c þ ðz z0Þ 2 ! p2 p2 þ ðz z0Þ 2 (2) where the parameters P, p, and pc are the material/probe interaction parameters16; the subscripts in brackets locate whether the parameter refers to an alumina/tetragonal zirconia mixture or to the monoclinic zirconia phase (A1T and M, respectively). Solving the integral Eq. (1) for VM(z) involves the knowledge of the mathematical character of the in-depth profile function. Based on previous Raman assessments from the sample cross sections in monolithic zirconia ceramics, we assume this function to be structured as follows16: VMð Þffi z A expð Þþ Bz Ctn ð Þ z þ 1 m þ V0 (3) where A, B, C, m, and n are constants, and V0 is the (initial) volume fraction of monoclinic phase in the bulk of the asreceived material. After substituting for Eq. (3) and the probe parameters from Table II in Eq. (1), an iterative computational routine was run using commercially available software24 to find the parameters A, B, C, m, and n, which best fit the experimental VM(z0) values, as retrieved from the in-depth confocal line scan. The results of this mathematical deconvolution procedure are plotted in Fig. 6. The deconvoluted monoclinic fraction profile, VM(z), is also compared in Fig. 6 with the monoclinic fraction profile typically found in the as-received material. This latter profile was almost unaltered after 10-h exposure in moist environment. On the other hand, after exposures in autoclave as long as 300 h, the thickness of material affected by an enhanced amount of monoclinic phase increased. This spectroscopic assessment may be invoked to explain the reason why the bulk toughness value was almost independent of autoclaving exposure time. In fact, we notice that, whatever steep the variance in the transformed near-surface portion of material, it remained confined to the level of a few tens of micrometers even after very long-term exposure in autoclave. The thickness of the transformation layer is important because it might affect the residual stress fields developed on the bearing surface and, thus, the wear behavior of the material. It has been reported30 that a compressive residual stress on the bearing surface contributes to reduce both amounts of wear and friction in alumina materials. In Pezzotti et al.,15 we have quantitatively measured the residual stress in the alumina phase on the surface of the composite before and after exposure to moist environment. A trend was found showing an initial moderate decrease to reach a magnitude of few tens of MPa in compression after about 20-h exposure (transformation thickness of about 6 mm) and then a gradual increase of the residual stress to reach zero magnitude after about 45 h and toward the tensile side (with aging time 445 h). This may suggest that the alumina matrix in the present composite could be more wear resistant than monolithic alumina. V. Conclusion In this study, we have applied combined techniques of fracture mechanics and confocal Raman microprobe spectroscopy to characterize both bulk and surface toughness of an alumina/zirconia composite used in advanced joint prostheses (e.g., BIOLOXs delta femoral heads). In a previous paper,15 we showed that transformation occurs in the material as a consequence of an environmentally assisted process. The autoclave testing procedure adopted in that study was quite severe and corresponded to exposures in human body of several lifetimes. The effect of an exposure of 10 h in autoclave at 1341C was also examined which, according to the activation energy for environmental resistance of the present composite, corresponds to a lifetime expectation of 39 years in vivo. Based on these results, the main outcomes of the present fracture mechanics assessments can be summarized as follows: (i) the crack-tip toughness level measured in the as-received material is comprehensive of a tangible contribution by transformation toughening; (ii) even after the material is environmentally aged for periods of the order of hundreds of hours and its surface toughness is reduced by about one-third, the measured toughness of the composite measured by any method is at least the same as that of monolithic alumina; (iii) the observed decrease of fracture toughness by about one-third is limited to the very surface of the material (i.e., to a layer of the order of a few tens of micrometers) and does not affect the bulk fracture toughness of the composite, as shown by Fig. 6. Profiles of monoclinic phase fraction, VM(z0), as collected with an in-depth confocal scanning along the subsurface of the sample autoclaved for 300 h at 1211C. The best-fit convoluted curves and the corresponding deconvoluted profiles (calculated according to Eqs. (1)–(3)) are also shown. Table II. Optical Parameters for Different Component Phases Crystal phase p (mm1 ) a (mm1 ) Alumina1tetragonal mixture 15.45 0.12 Monoclinic zirconia 4 0.03 August 2009 Advanced Ceramic Composite for Hip Prosthesis 1821
1822 Journal of the American Ceramic Society-Pezzonti et al. Vol. 92. No 8 the conspicuous invariance of SEVNB toughness values before 2N. J. Claussen. "Fracture Toughness of Al-O, with an Unstable ZO,Dis- and after long-term exposure in autoclave: (iv) microscopic Raman mapping at crack paths enabled us to qualitatively vi- umina Ceramics in Aqueous Solutions with Different pH, Wear, 254, 1141-6 sualize the transformation zone around the tip of an equilibrium (2003). crack and to clarify the contribution of phase transformation to Is. Novak and M. Kalin, "The Effect of ph on the Wear of Water-Lubricated toughness; and(v)care should be taken in directly comparing akakura, and R. P. Pitto, ""Raman Spectroscopic fracture toughness data collected by different methods(and even Analysis of Advanced Ceramic Composite for Hip Prosthesis,"J.Am. Ceram by the same method but with different testing parameters). A Soc.91向1199206(2008 rationale was offered to explain the discrepancy in toughness 16G. Pezzotti. T. Tateiwa. W. Zhu. T Kumakura and K. Yamamoto."Fluo- alues among different testing approaches ce Residual Stress Field Alumina Hip Joints, "J. Biomed. Opt, 11, 24009-18(2006) ng of Load-Bear- ial Joints. Ex Acknowledgments ial Standard, JIS R 1607. Japanese The authors are indebted to Prof. N. Sugano, MD, at Osaka University for his ai. ""Stable Crack Extension of an Alumina Ceramic in se),J. Soc. Mater. Sci.Jpv 42, 412-(199 wide Range Stress Intensity Fa for AsTm References tandard Fracture Toughness Specimens Wolfram, MATHEMATICA 4. Wolfram Research Inc, Champaign, IL M. Hamadouche. P. Boutin, J. Daussange. M. E. Bolander, and L. SedeL. J. Seidel and J. Rodel, Measurement of Crack Tip Toughness in Alumin llo. M. M Prospective Randomized Study, Clin. Orthop. Relat. Res, 436, 164-71(2005). 23. Kruzic and O Ritchie, " Determining the Toughness of Ceramics from Vic- m是1m0Am, Study am ce8图1 03)Pau ens. An penna ekow and v. P. Th ng Term Clinical Success 23. Chevalier. S. Deville. G. J. F. Bartolome. C. an. J. S. of Advanced Ceramic Prostheses, J. Mater. Sci. Mater. Med., 18[1] 47-56 Moya, L. A Diaz, and R. Torrecillas. ""Nanostructured Ceramic Oxides with T. Pulliam and R. T. Trousdale. "Fracture of a Ceramic Femoral Head sl9n-Glock sowth Resistance Close to Covalent Materials " Nano Let. 5. After a Revision Operation. A Case Report. J. Bone Jt. Surg. Am. 79. 118-9 enzaid, J. Chevalier, M. Saadaoui. G. Fantozzi, M. Na and R. Torrecillas. ""Fracture Toughness, Strength and Slow Crack Growth in a eY -C Ha. S.-Y. Kim. H. J. Kim. J. J. Yoo and K-H. Koo"Ceramic Liner De Aza. J. Chevalier. G. Fantozz. M. Schehl and R. torrecilla G. Ya Akimov and V M. Timchenko. " Effect of Strain Rate and preloading G. Pfaff. R. Rack, and G. willmann. " Long-Term Performance of a T on the Fracture Toughness of ZrO2-Based Ceramics. Strength Mater, 34[5151 formation Toughened Platelet Reinforced Alumina Matrix Composite, Bioce- mics,14,541-2(2001) Pezzotti, Y. Okamoto. T. Nishida, and M. Sakai."On the Near-Ti alidation of a New High Performance posite for Use in Total Joint Repla g in Particulate and Platelet- Reinforced ramics, "Acta Mater, 44[3]899-914(1996) IG. Insleyand and R Teicher, "Next Generation C vak, M. Kalin, P. Lukas. G. Anne, J. Vleugels and O. van Der Biest Toughened Alumina for Hip Joint Prostheses, Key En bosites on Their Wear and Friction Behaviour. J. Eur. Ceram. Soc. 27. 151--6 M. Ruhle. A. H. Heuer and N. Claussen. ""Transformation and microcrack 2007) Toughening omplementary Processes in Zirconia-Toughened Alumina. J.Am. Ceran.Soc,69,195-7(1986
the conspicuous invariance of SEVNB toughness values before and after long-term exposure in autoclave; (iv) microscopic Raman mapping at crack paths enabled us to qualitatively visualize the transformation zone around the tip of an equilibrium crack and to clarify the contribution of phase transformation to toughness; and (v) care should be taken in directly comparing fracture toughness data collected by different methods (and even by the same method but with different testing parameters). A rationale was offered to explain the discrepancy in toughness values among different testing approaches. Acknowledgments The authors are indebted to Prof. N. Sugano, MD, at Osaka University for his helpful discussions. References 1 M. Hamadouche, P. Boutin, J. Daussange, M. E. Bolander, and L. Sedel, ‘‘Alumina-on-Alumina Total Hip Arthroplasty: A Minimum 18.5-Year FollowUp Study,’’ J. Bone Jt. Surg. Am., 84-A [1] 69–77 (2002). 2 J. D’Antonio, W. Capello, M. Manley, M. Naughton, and K. Sutton, ‘‘Alumina Ceramic Bearings for Total Hip Arthroplasty: Five-Years Results of a Prospective Randomized Study,’’ Clin. Orthop. Relat. Res., 436, 164–71 (2005). 3 L. Sedel, ‘‘Evolution of Alumina-on-Alumina Implants: A Review,’’ Clin. Orthop., 379, 8–54 (2000). 4 D. Rekow and V. P. Thompson, ‘‘Engineering Long Term Clinical Success of Advanced Ceramic Prostheses,’’ J. Mater. Sci. Mater. Med., 18 [1] 47–56 (2007). 5 I. T. Pulliam and R. T. Trousdale, ‘‘Fracture of a Ceramic Femoral Head After a Revision Operation. A Case Report,’’ J. Bone Jt. Surg. Am., 79, 118–9 (1997). 6 Y.-C. Ha, S.-Y. Kim, H. J. Kim, J. J. Yoo, and K.-H. Koo, ‘‘Ceramic Liner Fracture after Cementless Alumina-on-Alumina Total Hip Arthroplasty,’’ Clin. Orthop. Relat. Res., 458, 106–10 (2006). 7 W. Burger and H. G. Richter, ‘‘High Strength and Toughness Alumina Matrix Composites by Transformation Toughened and ‘In Situ’ Platelet Reinforcement (ZPTA)—The New Generation of Bioceramics,’’ Bioceramics, 13, 545–8 (2000). 8 H.-G. Pfaff, R. Rack, and G. Willmann, ‘‘Long-Term Performance of a Transformation Toughened Platelet Reinforced Alumina Matrix Composite,’’ Bioceramics, 14, 541–2 (2001). 9 M. Kuntz, ‘‘Validation of a New High Performance Alumina Matrix Composite for Use in Total Joint Replacement,’’ Semin. Arthroplasty, 17, 141–5 (2006). 10G. Insleyand and R. Steicher, ‘‘Next Generation Ceramics Based on Zirconia Toughened Alumina for Hip Joint Prostheses,’’ Key Eng. Mater., 254–6, 675–8 (2004). 11M. Ru¨hle, A. H. Heuer, and N. Claussen, ‘‘Transformation and Microcrack Toughening as Complementary Processes in Zirconia-Toughened Alumina,’’ J. Am. Ceram. Soc., 69, 195–7 (1986). 12N. J. Claussen, ‘‘Fracture Toughness of Al2O3 with an Unstable ZrO2 Dispersed Second Phase,’’ J. Am. Ceram. Soc., 61, 49–51 (1976). 13M. Kalin, S. Novak, and J. Vizinitin, ‘‘Wear and Friction Behavior of Alumina Ceramics in Aqueous Solutions with Different pH,’’ Wear, 254, 1141–6 (2003). 14S. Novak and M. Kalin, ‘‘The Effect of pH on the Wear of Water-Lubricated Alumina and Zirconia Ceramics,’’ Tribol. Lett., 17, 727–32 (2004). 15G. Pezzotti, K. Yamada, S. Sakakura, and R. P. Pitto, ‘‘Raman Spectroscopic Analysis of Advanced Ceramic Composite for Hip Prosthesis,’’ J. Am. Ceram. Soc., 91 [4] 1199–206 (2008). 16G. Pezzotti, T. Tateiwa, W. Zhu, T. Kumakura, and K. Yamamoto, ‘‘Fluorescence Spectroscopic Analysis of Surface and Sub-Surface Residual Stress Fields in Alumina Hip Joints,’’ J. Biomed. Opt., 11, 24009–18 (2006). 17G. Pezzotti, ‘‘Stress Microscopy and Confocal Raman Imaging of Load-Bearing Surfaces in Artificial Joints,’’ Expert Rev. Med. Devices, 4 [2] 165–89 (2007). 18Japanese Standards Association. Testing Methods for Fracture Toughness on High Performance Ceramics. Japan Industrial Standard, JIS R 1607. Japanese Standards Association, Tokyo, Japan, 1990. 19T. Nojima and O. Nakai, ‘‘Stable Crack Extension of an Alumina Ceramic in Three Point Bending Test (in Japanese),’’ J. Soc. Mater. Sci. Jpn., 42, 412–8 (1993). 20J. E. Srawley, ‘‘Wide Range Stress Intensity Factor Expressions for ASTM E399 Standard Fracture Toughness Specimens,’’ Int. J. Fract., 12 [3] 475–6 (1976). 21S. Wolfram, MATHEMATICA 4. Wolfram Research Inc., Champaign, IL, 2000. 22J. Seidel and J. Ro¨del, ‘‘Measurement of Crack Tip Toughness in Alumina as a Function of Grain Size,’’ J. Am. Ceram. Soc., 80 [2] 433–8 (1997). 23T. Fett, A. B. Kounga, and J. Ro¨del, ‘‘Crack Opening Displacements of Vickers Indentation Cracks,’’ Eng. Fract. Mech., 72, 647–59 (2005). 24J. Kruzic and O. Ritchie, ‘‘Determining the Toughness of Ceramics from Vickers Indentations Using the Crack-Opening Displacements: An Experimental Study,’’ J. Am. Ceram. Soc., 86 [8] 1433–6 (2003). 25J. Chevalier, S. Deville, G. Fantozzi, J. F. Bartolome´, C. Pecharroman, J. S. Moya, L. A. Diaz, and R. Torrecillas, ‘‘Nanostructured Ceramic Oxides with a Slow Crack Growth Resistance Close to Covalent Materials,’’ Nano Lett., 5 [7] 1297–310 (2005). 26R. Benzaid, J. Chevalier, M. Saaˆdaoui, G. Fantozzi, M. Nawa, L. A. Diaz, and R. Torrecillas, ‘‘Fracture Toughness, Strength and Slow Crack Growth in a Ceria Stabilized Zirconia–Alumina Nanocomposite for Medical Applications,’’ Biomaterials, 29 [27] 3636–41 (2008). 27A. H. De Aza, J. Chevalier, G. Fantozzi, M. Schehl, and R. Torrecillas, ‘‘Crack Growth Resistance of Alumina, Zirconia and Zirconia Toughened Alumina Ceramics for Joint Prostheses,’’ Biomaterials, 23 [3] 937–45 (2002). 28G. Ya. Akimov and V. M. Timchenko, ‘‘Effect of Strain Rate and Preloading on the Fracture Toughness of ZrO2-Based Ceramics,’’ Strength Mater., 34 [5] 514– 7 (2002). 29G. Pezzotti, Y. Okamoto, T. Nishida, and M. Sakai, ‘‘On the Near-Tip Toughening by Crack-Face Bridging in Particulate and Platelet-Reinforced Ceramics,’’ Acta Mater., 44 [3] 899–914 (1996). 30S. Novak, M. Kalin, P. Lukas, G. Anne, J. Vleugels, and O. Van Der Biest, ‘‘The Effect of Residual Stresses in Functionally Graded Alumina–ZTA Composites on Their Wear and Friction Behaviour,’’ J. Eur. Ceram. Soc., 27, 151–6 (2007). & 1822 Journal of the American Ceramic Society—Pezzotti et al. Vol. 92, No. 8