M.B. Ruggles-Wrenn, N.R. Szymczak/Composites: Part A 39(2008)1829-1837 several non-oxide Cmcs exhibited significant susceptibility at tem- Combining Eqs. (5)and (3), creep lifetime can be expressed in terms peratures >1100C, with the n values ranging from 6 to 20. of the crack growth parameters obtained from constant stress-rate Based on experimental results, Choi and Bansal 29 determined data as: that for a 2-D Hi-Nicalon-fiber-reinforced CMC at 1100C the gov- D+1 low crack growth. A power-law slow crack growth model for t(n+ion rning failure mechanism in shear was also associated with the hear was proposed and formulated to account for the dependence of shear strength of the composite on loading rate. The equation where tr and o are time to failure and applied creep stress, giving shear strength function of applied shear stress-rate respectively SCG parameters in shear, ns and D. were determined from composite at 1200C in steam [38]. Excellent agreement betwee 7 sas identical in form to Eq (2)established for tensile loading. Eq(6)was used to predict the tensile creep lifetimes of N720/ the slope and intercept, respectively, of the linear regression anal- the predictions and the experimental results indicated that the ysis of the log(shear strength)vs log(applied stress-rate) plot of environmentally assisted slow crack growth was indeed the gov the experimental data. The SCG parameters for interlaminar shear erning failure mechanism for N720/A CMC at 1200C in steam. were ns-11.2+2. 2 and Ds-1924*0.91. The SCG parameters for Choi et al. [27, 28] demonstrated that for various CMCs at 1100 in-plane shear were determined as ns=11.4+1.9 andand 1200C the predictions of tensile creep lifetimes based on D=33.27±1.35 he constant stress-rate test data were in good agreement with In the case of glass and ceramic materials that exhibit subcriti- the experimental results. Moreover, Choi and Bansal [29] showed cal(slow) crack growth as the predominant failure mechanism, that creep lifetime in shear can be successfully predicted by using time to failure under constant stress can be predicted from con- Eq (6 )where the parameters n and d are replaced with the SCG stant stress-rate test data by using the linear elastic crack growth parameters in shear ns and Ds, and applied tensile stress o is re- model [27, 28]. Using the empirical power-law crack-velocity for- placed with applied shear stress t mulation in Eq (1)the time to failure under constant stress can It is recognized that the compressive failure in fiber-reinforced be derived in the form (411 com s generally associated with micro- buckling of the fibers 42-45]. Flexural stresses in a fiber due to in- 2 rico (5) phase buckling lead to the formation of kink zones, which can AY(n-2) cause fracture in brittle fibers [46, 47]. In the case of the 0/90 18 mm 18 mm ig. 8. Fracture surfaces of N720/A specimens tested in compression with the stress-rate of 25 MPa/s at 1200 C:(a)in air and(b)in steam.several non-oxide CMCs exhibited significant susceptibility at tem￾peratures P1100 C, with the n values ranging from 6 to 20. Based on experimental results, Choi and Bansal [29] determined that for a 2-D Hi-Nicalon-fiber-reinforced CMC at 1100 C the gov￾erning failure mechanism in shear was also associated with the slow crack growth. A power-law slow crack growth model for shear was proposed and formulated to account for the dependence of shear strength of the composite on loading rate. The equation giving shear strength as a function of applied shear stress-rate was identical in form to Eq. (2) established for tensile loading. The SCG parameters in shear, ns and Ds, were determined from the slope and intercept, respectively, of the linear regression anal￾ysis of the log (shear strength) vs log (applied stress-rate) plot of the experimental data. The SCG parameters for interlaminar shear were ns = 11.2 ± 2.2 and Ds = 19.24 ± 0.91. The SCG parameters for in-plane shear were determined as ns = 11.4 ± 1.9 and Ds = 33.27 ± 1.35. In the case of glass and ceramic materials that exhibit subcriti￾cal (slow) crack growth as the predominant failure mechanism, time to failure under constant stress can be predicted from con￾stant stress-rate test data by using the linear elastic crack growth model [27,28]. Using the empirical power-law crack-velocity for￾mulation in Eq. (1) the time to failure under constant stress can be derived in the form [41]: tf ¼ 2 K2 ICrn2 i AY2 ðn 2Þ " #rn ð5Þ Combining Eqs. (5) and (3), creep lifetime can be expressed in terms of the crack growth parameters obtained from constant stress-rate data as: tf ¼ Dnþ1 ðn þ 1Þ " #rn ð6Þ where tf and r are time to failure and applied creep stress, respectively. Eq. (6) was used to predict the tensile creep lifetimes of N720/A composite at 1200 C in steam [38]. Excellent agreement between the predictions and the experimental results indicated that the environmentally assisted slow crack growth was indeed the gov￾erning failure mechanism for N720/A CMC at 1200 C in steam. Choi et al. [27,28] demonstrated that for various CMCs at 1100 and 1200 C the predictions of tensile creep lifetimes based on the constant stress-rate test data were in good agreement with the experimental results. Moreover, Choi and Bansal [29] showed that creep lifetime in shear can be successfully predicted by using Eq. (6) where the parameters n and D are replaced with the SCG parameters in shear ns and Ds, and applied tensile stress r is re￾placed with applied shear stress s. It is recognized that the compressive failure in fiber-reinforced composites is a complex process generally associated with micro￾buckling of the fibers [42–45]. Flexural stresses in a fiber due to in￾phase buckling lead to the formation of kink zones, which can cause fracture in brittle fibers [46,47]. In the case of the 0/90 Fig. 8. Fracture surfaces of N720/A specimens tested in compression with the stress-rate of 25 MPa/s at 1200 C: (a) in air and (b) in steam. 1834 M.B. Ruggles-Wrenn, N.R. Szymczak / Composites: Part A 39 (2008) 1829–1837