degradation with decreasing stress rate has been known to be due to slow crack growth (delayed failure or fatigue) of an initial crack, typically governed by the following empirical power-law relation [1-3] n I IC v A K K ) ( / = (1) where v, KI and KIC are crack velocity, mode I stress intensity factor and fracture toughness, respectively. A and n are called slow crack growth (SCG) parameters. Based on this power-law relation, the strength (sf) can be derived as a function of applied stress rate (s& ) [1-3], 1 1 [ ] + = n f D s s & (2) where D is another SCG parameter associated with inert strength, n and crack geometry. Equation (2) can be expressed in a more convenient form by taking logarithms of both sides D n f log log 1 1 log + + = s s & (3) Constant stress-rate (“dynamic fatigue”) testing based on Equation (2) or (3) has been established as ASTM Test Methods (C1368 [2] and C1465 [3]) to determine SCG parameters of advanced monolithic ceramics at ambient and elevated temperatures. It has been recommended to use units of MPa for sf and MPa/s for s& [23]. As can be seen in Figure 1, the data fit to Equation (3) is very reasonable with the coefficients of correlation (rcoef) all greater than 0.980, indicating that the damage evolution/accumulation or delayed failure of the composite materials would be adequately described by the power-law type relation, Equation (1). Assuming this, the apparent parameters n’ and D’ for the composites were determined using a linear regression analysis based on Equation (3) with the data in Figure 1. Values of n’ = 9.0 and D’ = 226, n’ = 18 and D’ = 185, and n’ = 6 and D’ = 158 were obtained for SiC/CAS-II, SiC/MAS-5 and SiC/SiC, respectively (The prime was used here for composite materials to distinguish them from monolithic ceramic counterparts.). It is noteworthy that the value of n’, a measure of susceptibility to damage, was very low for both SiC/CAS-II and SiC/SiC, but intermediate for SiC/MAS-5. Typical monolithic silicon nitrides and silicon carbides at high temperatures at ³1200oC exhibit n ³ 20. Hence, compared with monolithic ceramics, the SiC/CAS-II and SiC/SiC composites exhibited a significantly higher susceptibility to damage evolution/accumulation. The preloading or accelerated testing technique has been developed for monolithic ceramics in order to save test time in constant stress-rate testing [8]. Based on the power-law SCG relation of Equation (1) with some mathematical manipulation, strength of a test specimen under a preload (ap) was derived as a function of preloading factor as follows [8,2,3]: 1 1 1) (1 + + + = n n p f fp s s a (4) where sfp is strength with a preload and ap (0£ ap£1) is a preloading factor (or percentage of preload) in which a preload stress (applied to the test specimen) is normalized with respect to the strength with zero preload. Equation (4) indicates that strength with a preload is sensitive to the magnitude of preload particularly at lower n and higher ap values. A theoretical prediction of ultimate strength as a function of preload, based on Equation (4) with estimated values of n’ from Figure 1, is presented in Figure 2. The prediction is in excellent agreement with the experimental data for all the three composite materials tested, as seen in the figure. This result obtained from the composite materials is also analogous to that observed in advanced monolithic ceramics and glass [8]. Damage, mainly SCG, of monolithic ceramics occurs substantially close to 90 % of total failure time because of their higher n (³20) value [8]. The applicability of the preloading analysis for the composite materials strongly suggests that major damage evolution/accumulation process would be the one governed by the power-law relation (Equation (1)) and that the damage would have occurred after a long incubation time, at least after 80 % of total test time. The strength dependency on test rate, the very reasonable data fit to Equation (3) and the applicability of preloading technique all support that the damage evolution/accumulation of the composite materials tested was controlled by a process very similar, in principle, to the power-law type of SCG of monolithic ceramics.
RkJQdWJsaXNoZXIy MjM0NDE=