G. de Portu et al Acta Materialia 53(2005)1511-1520 Wavenumber(cm) Wavenumber(cm 460cm1 =2600005cmGP =-140:±0.15 cm-I/GPa Stress(MPa) Fig. 1. Typical recorded for Al2O3(fluorescence)and 3Y-TZP(Raman) are shown in(a) and(b), respectively. Results of uniaxial stress in(c)and(d) for selected bands. The correspondent piezo spectroscopic coefficients Iu= l; 3 are shown together with the recorded data tent, especially at intermediate volume fractions. This means that not only the Cte mismatch between AlO3 and ZrO2 phase (which produces the spectral shift shown in Fig. 2(a)), but also the elastic mismatch be- tween these two phases(responsible for the dependence of li on Al2O3 volume fraction as shown in Fig. 2(b)) should be taken into account for a reliable residual stress evaluation in composite materials. Using the two Al,O3 volume fraction(%) master curves obtained from this set of calibrations (Fig. 2(a)and(b)), for an AlO3/ZrO2 composite con- taining a known volume fraction of Al_O3 dispersoid the"zero-stress"frequency and Ii value can be system atically obtained. It should be noted that average fre- quencies and piezo-spectroscopic matrix traces as a function of Al,O3 volume fraction(Fig. 2(a) and (b), respectively)can be regarded as values proper to uni- formly distributed composite materials, provided that the volume screened by the piezo-spectroscopic probe in each stress measurement is sufficiently large to include Al2O3 volume fraction(%) a statistically meaningful number of Al2O3 dispersoids Fig. 2. Dependence of the Ri band position of externally unstressed Using the"zero-stress "frequency and a li values from chromophore Al203 as a function of its volume fraction when Fig. 2(a)and(b), respectively, for obtaining stress mbedded in a 3Y-TZP matrix (a). Piezo-spectroscopic coefficient assessments through Eq. (1), the effect of the microstress Iu= 1i3 as recorded for the Ri band for specimens in(a) field on the experimentally determined spectral shift of the Al,O3 phase can be automatically subtracted from grain-to-grain residual stresses between Al2O, and 3Y- the net measured shift. In other words, the Al2O3 phase TZP phases. Fig. 2(b)represents the dependence on uni- can be used as a"stress sensor", because the shift of its axial stress of the trace of the piezo-spectroscopic ma- spectral lines directly reads the macroscopic stress field trix, Ili, for the R band of Al2O3, when Al,O3 fine piled up among different layers of the structure. In this powder was incorporated with different volume frac- context, it should be noted that, in a previous study of tions into sintered 3Y-TZP. As seen, the variation of residual stresses in AlO3/ZrO, laminates [6], the"stress Iii with volume fraction of AlO3 is of a significant ex- sensor"approach was similarly adopted to evaluate thegrain-to-grain residual stresses between Al2O3 and 3Y￾TZP phases. Fig. 2(b) represents the dependence on uni￾axial stress of the trace of the piezo-spectroscopic ma￾trix, Pii, for the R1 band of Al2O3, when Al2O3 fine powder was incorporated with different volume frac￾tions into sintered 3Y-TZP. As seen, the variation of Pii with volume fraction of Al2O3 is of a significant ex￾tent, especially at intermediate volume fractions. This means that not only the CTE mismatch between Al2O3 and ZrO2 phase (which produces the spectral shift shown in Fig. 2(a)), but also the elastic mismatch be￾tween these two phases (responsible for the dependence of Pii on Al2O3 volume fraction as shown in Fig. 2(b)) should be taken into account for a reliable residual stress evaluation in composite materials. Using the two master curves obtained from this set of calibrations (Fig. 2(a) and (b)), for an Al2O3/ZrO2 composite con￾taining a known volume fraction of Al2O3 dispersoid, the ‘‘zero-stress’’ frequency and Pii value can be system￾atically obtained. It should be noted that average fre￾quencies and piezo-spectroscopic matrix traces as a function of Al2O3 volume fraction (Fig. 2(a) and (b), respectively) can be regarded as values proper to uni￾formly distributed composite materials, provided that the volume screened by the piezo-spectroscopic probe in each stress measurement is sufficiently large to include a statistically meaningful number of Al2O3 dispersoids. Using the ‘‘zero-stress’’ frequency and a Pii values from Fig. 2(a) and (b), respectively, for obtaining stress assessments through Eq. (1), the effect of the microstress field on the experimentally determined spectral shift of the Al2O3 phase can be automatically subtracted from the net measured shift. In other words, the Al2O3 phase can be used as a ‘‘stress sensor’’, because the shift of its spectral lines directly reads the macroscopic stress field piled up among different layers of the structure. In this context, it should be noted that, in a previous study of residual stresses in Al2O3/ZrO2 laminates [6], the ‘‘stress sensor’’ approach was similarly adopted to evaluate the 200 400 600 Wavenumber (cm-1) Intensity (a.u.) 460 cm-1 -200 -100 0 100 200 -1 0 1 Stress (MPa) Peak shift (cm-1) 460 cm-1 Πu = -1.40:± 0.15 cm-1/GPa Peak shift (cm-1) -200 -100 0 100 200 -1 0 1 Stress (MPa) R1 Πu = -2.60:± 0.0025 cm-1/GPa 6000 6050 6100 6150 Intensity (a.u.) Wavenumber (cm-1) R1 R2 (a) (b) (c) (d) Fig. 1. Typical spectra recorded for Al2O3 (fluorescence) and 3Y-TZP (Raman) are shown in (a) and (b), respectively. Results of uniaxial stress calibrations are shown in (c) and (d) for selected bands. The correspondent piezo-spectroscopic coefficients Pu = Pii/3 are shown together with the recorded data scatter. 0 20 40 60 80 100 6085 6086 6087 6088 6089 6090 Al2O3 volume fraction (%) Zero-stress position (cm-1) Al2O3 volume fraction (%) PS coefficient (cm-1/GPa) 0 20 40 60 80 100 -3 -2 (a) (b) Fig. 2. Dependence of the R1 band position of externally unstressed chromophore Al2O3 as a function of its volume fraction when embedded in a 3Y-TZP matrix (a). Piezo-spectroscopic coefficient Pu = Pii/3 as recorded for the R1 band for specimens in (a). 1514 G. de Portu et al. / Acta Materialia 53 (2005) 1511–1520