Plane strain crack growth models for fatigue crack growth life predictions

Experimental data and analytical models have shown that a growing fatigue crack produces a plastic wake. This, in turn, leads to residual compressive stresses acting over the crack faces during the unloading portion of the fatigue cycle. This crack closure effect results in an applied stress intensity factor during unloading which is greater than that associated with the K-min, thus producing a crack-driving force which is less than Delta K = K-max - K-min. Life predictions which do not account for this crack closure effect give inaccurate life estimates, especially for fully reversed loadings. This paper discusses the development of a crack closure expression for the 4-point bend specimen using numerical results obtained from a modified strip-yield model. Data from tests of eight 4-point bend specimens were used to estimate the specimen constraint factor (stress triaxiality effect). The constraint factor was then used in the estimation of the crack opening stresses for each of the bend rests. The numerically estimated crack opening stresses were used to develop an effective stress intensity factor range, Delta K-eff. The resulting crack growth rate data when plotted versus Delta K-eff resulted in a material fatigue crack growth rate property curve independent of test specimen type, stress level, and R-ratio. Fatigue crack growth rate data from center-cracked panels using Newman's crack closure model, from compact specimens using Eason's R-ratio expression, and from bend specimens using the model discussed in this paper are all shown to fall along the same straight line (on log-log paper) when plotted versus Delta K-eff, even though crack closure differs for each specimen type.