CRACK-TIP PARAMETERS FOR CREEP-BRITTLE CRACK-GROWTH

Crack growth histories for creep-brittle aluminium alloy 2519-T87 are simulated by controlling the rate of release of finite element nodes a long the crack growth path using a variable time-step, nodal release a lgorithm. While earlier experimental studies established little or no correlation between time-dependent fracture parameters and the crack g rowth rate, (a) over dot, during creep-brittle fracture, the numerical results presented here indicate Excellent correlation of the C-t para meter with (a) over dot during the quasi-steady state crack growth reg ime. Differences in the experimental and numerical determination of ti me-dependent fracture parameters are likely to be due to difficulties in experimental determination of the creep component of the load line deflection rate, (V) over dot (c), during creep-brittle crack growth. A new quantity, K/r(c)(q), is derived from time-dependent fracture par ameters to predict crack growth for transient, quasi-steady state and steady state crack growth. However, C-t and K/r(c)(q) should only be e mployed as parameters for predicting creep-brittle crack growth with a n understanding of the couplings which exist between these parameters and the crack growth rate.