3250 H. Mei, L. Cheng Materials Letters 59(2005)3246-3251 101÷ obtained by electrical resistance measurement on the basis the newly-developed SDIA. Fig. 5 dem 85≌5 correlation between vibrations of resistance R(i)(calculation according to formula (1) of the specimen and cyclic temperature during thermal cycle testing. Real time resis tances change periodically with thermal cycles and both H periods of them are equal approximately(about 120 s). It is more interesting that resistance decreases upon heating and 0170018001900 increases with cooling in each cycle. As expected, the greater the damage in cooling, the more severe are the Time(S) spikes or resistance increase. It is not surprised that the Fig. 5. Correlation between change in resistance(left) acquired by SDIA resistance of the composites is rather larger in cooling than and cyclic temperature(right) in testing. upon heating because fibers are broken and parted from each other once cooling. In addition, as previously described, the damage in the initial several cycles is had enough space to tolerate micro-thermal-expansion. The considerable severe due to the rapid physical destruction. value of transverse axis parallel to inflexion of the curve is Incremental damage, however, diminished upon thermal about 38 times. After this point, the continuous and slow decrease in mechanical properties should be ascribed to cycling. A typical relative resistance change (i.e, a ratio of the r(i to the virgin resistance Ro)vs. thermal cycle curve fibers oxidation in wet oxygen with increasing therma of the specimen in the wet oxygen atmosphere is pre cycling time. Accordingly, thermal cycling damage to C/SiC in Fig. 6. The relative resistance also decreases reversibly composites is limited and there exists a critic cal cyclin upon heating and increases with cooling in every cycle. As cycling progresses, the baseline resistance decreases con- tinuously and then levels off after about 35 times, which is 3.3. In situ monitoring of damage by SDIA during the consistent with the above experimental statistical results thermal cycling test (about 38 cycles according to Fig. 4). During the initial stage of thermal cycle testing, with increasing number, The damage revealed by electrical resistance measure- surface cracks propagate inwards, which lead to an increase ment is subtle, in contrast to damage in the form of well- defined delamination cracks, fiber breakage or del of crack density in composites. The increase of matrix crack bonded density results in the rapidly decrease of the mechanical regions in composites. Since carbon fibers were electrical properties of composites by increasing broken fibers conductors(p=2.0 x 10>.m), the measurement of the According to the left curve in Fig. 6, the energy absorbed variations of electrical resistance appears to be a valuable into specimen is very high and rapidly dissipated to destroy technique for in situ monitoring internal damage evolution the materials at the first stage. Thermal cycling damage, at of the composites. In the case of a cross-ply [0/90] carbon this time, is a dominant factor. Subsequently, relative minor tiber SiC-matrix composite specimen, conductivity depends damage occurs gradually as cycling progresses, leading to on the damage degree of (o%) continuous fibers parallel to the gradual and irreversible decrease of the baseline the longitudinal direction. As shown in Fig 3, two types of resistance. The accumulated resistance of specimen(Rtotal) major fibers failure pattern are observed: (i) physical fracture under thermal cycling and (ii) chemical recession is a mathematic integral of the left curves according to formula(2)(i.., the area under the left curves). The right in oxidizing atmosphere. The failure of carbon fiber reinforcements can directly lead to decrease in mechanical properties and increase in resistance. The changes in resistance, thus, correlate with damage of the composite The resistance of the composite R, may be written as follows [3], 10三 7d=4 2+4Py (6) where Pr is the specific electrical resistivity of carbon fiber, Vr the volume fraction of unbroken fibers, L the length between the electrodes on both ends of the specimen, b and d specimen width and thickness, respectively, and Re the 010203040506 contact resistance between the sample and the electrodes In the present work, the law of internal damage in Fig. 6. R(/Ro vs thermal cycles(left) recorded by SDIA and accumulated composites during the dynamic thermal cycle testing can be resistance of specimen(right)in testinghad enough space to tolerate micro-thermal-expansion. The value of transverse axis parallel to inflexion of the curve is about 38 times. After this point, the continuous and slow decrease in mechanical properties should be ascribed to fibers oxidation in wet oxygen with increasing thermal cycling time. Accordingly, thermal cycling damage to C/SiC composites is limited and there exists a critical cycling number. 3.3. In situ monitoring of damage by SDIA during the thermal cycling test The damage revealed by electrical resistance measure￾ment is subtle, in contrast to damage in the form of well￾defined delamination cracks, fiber breakage or debonded regions in composites. Since carbon fibers were electrical conductors (q = 2.0105 V&m), the measurement of the variations of electrical resistance appears to be a valuable technique for in situ monitoring internal damage evolution of the composites. In the case of a cross-ply [0/90-] carbon fiber SiC-matrix composite specimen, conductivity depends on the damage degree of (0-) continuous fibers parallel to the longitudinal direction. As shown in Fig. 3, two types of major fibers failure pattern are observed: (i) physical fracture under thermal cycling and (ii) chemical recession in oxidizing atmosphere. The failure of carbon fiber reinforcements can directly lead to decrease in mechanical properties and increase in resistance. The changes in resistance, thus, correlate with damage of the composite. The resistance of the composite R, may be written as follows [3], R ¼ qfL bdVf þ Rc ð6Þ where qf is the specific electrical resistivity of carbon fiber, Vf the volume fraction of unbroken fibers, L the length between the electrodes on both ends of the specimen, b and d specimen width and thickness, respectively, and Rc the contact resistance between the sample and the electrodes. In the present work, the law of internal damage in composites during the dynamic thermal cycle testing can be obtained by electrical resistance measurement on the basis of the newly-developed SDIA. Fig. 5 demonstrates that a correlation between vibrations of resistance R(i) (calculation according to formula (1)) of the specimen and cyclic temperature during thermal cycle testing. Real time resis￾tances change periodically with thermal cycles and both periods of them are equal approximately (about 120 s). It is more interesting that resistance decreases upon heating and increases with cooling in each cycle. As expected, the greater the damage in cooling, the more severe are the spikes or resistance increase. It is not surprised that the resistance of the composites is rather larger in cooling than upon heating because fibers are broken and parted from each other once cooling. In addition, as previously described, the damage in the initial several cycles is considerable severe due to the rapid physical destruction. Incremental damage, however, diminished upon thermal cycling. A typical relative resistance change (i.e., a ratio of the R(i) to the virgin resistance R0) vs. thermal cycle curve of the specimen in the wet oxygen atmosphere is presented in Fig. 6. The relative resistance also decreases reversibly upon heating and increases with cooling in every cycle. As cycling progresses, the baseline resistance decreases con￾tinuously and then levels off after about 35 times, which is consistent with the above experimental statistical results (about 38 cycles according to Fig. 4). During the initial stage of thermal cycle testing, with increasing number, surface cracks propagate inwards, which lead to an increase of crack density in composites. The increase of matrix crack density results in the rapidly decrease of the mechanical properties of composites by increasing broken fibers. According to the left curve in Fig. 6, the energy absorbed into specimen is very high and rapidly dissipated to destroy the materials at the first stage. Thermal cycling damage, at this time, is a dominant factor. Subsequently, relative minor damage occurs gradually as cycling progresses, leading to the gradual and irreversible decrease of the baseline resistance. The accumulated resistance of specimen (Rtotal) is a mathematic integral of the left curves according to formula (2) (i.e., the area under the left curves). The right Change in resistance (%) 0 2 4 6 8 10 Time (S) 1600 1700 1800 1900 2000 2100 Temperature (°C) 700 1200 Fig. 5. Correlation between change in resistance (left) acquired by SDIA and cyclic temperature (right) in testing. Thermal Cycles 0 10 20 30 40 50 60 0 0.5 1.0 RTotal ( Ω) R(i)/R0 (%) 0 5 10 1.5 × 104 Fig. 6. R(i)/R0 vs. thermal cycles (left) recorded by SDIA and accumulated resistance of specimen (right) in testing. 3250 H. Mei, L. Cheng / Materials Letters 59 (2005) 3246 – 3251