FATIGUE CRACK GROWTH MEASUREMENTS AND PREDICTIONS Test Specimens The test specimens for measurements of fatigue crack growth were of 6 mm thickness with a central hole of radius 3 mm, a width of 100 mm and a length of 245 mm. Two initial starter cracks were machined using an EDM technique on one face of the specimen and on both sides of the hole [8]. For specimens that had been subjected to cold expansion the starter cracks were located on the entrance face, defined by the FTI method. Fatigue Loading The specimens were fatigue loaded in a 250 kN servo hydraulic test machine at a rate of 10 cycles per second. The maximum load applied to the specimens was equivalent to a far field stress of 162 MPa. Various R ratios were used, but in the results described here an R ratio of 0.1 was used. Surface crack growth was measured on both surfaces of the specimen and on both sides of the hole using vernier microscopes as a function of the number of cycles of load. From these measurements, the rate of crack growth versus crack length was calculated. Fatigue Properties The crack growth rate was assumed to be a function of the effective range of stress intensity factor. This function was measured using a specimen of dimensions defined above with a non cold expanded hole. To ensure the measurements were not effected by closure, a high R ration of 0.7 was used. Finite Element Predictions Finite element predictions of fatigue crack growth rate were made using a two dimensional model of the test specimen. The model was run repeatedly using different crack lengths and the stress intensity factor calculated for each length. Predictions were also made of the load required to cause the crack to open enabling the effective range of stress intensity factor to be calculated for each crack length. The function of crack growth rate versus stress intensity range derived form high R ratio tests was then used to make predictions of crack growth rate. Results Measurements of crack length versus number of cycles are shown in Figure 3. The crack length in the figure is the average of the two lengths measured on the entrance face of the specimen. The measurements for three cases are presented: for a non cold expanded hole, a cold expanded hole and a cold expanded hole with creep relaxation while under applied load. Cold expansion can be seen to lower substantially the rate of fatigue crack growth. Creep relaxation removes some but not all of the benefit of cold expansion. Figure 4 presents the experimental measurements of crack growth rate versus the crack length for the case of a specimen with a cold expanded hole. Measurements for both cracks on the entrance face of the specimen are provided: the crack to the left and to the right of the hole. Also shown are the finite element predictions of crack growth rate. The comparison shows the finite element model underpredicts the growth rate for small crack lengths and over-predicts the rate for large crack lengths. Because of the two dimensional nature of the finite element model, the crack is assumed to have a parallel through-the-thickness geometry whereas the actual crack has a complex three dimensional shape for small crack lengths. At large crack lengths the finite element predictions suggest the crack is always open during the entire load cycle whereas substantial closure is observed in the experiment. This closure is believed to be due to plasticity around the crack tip and would tend to reduce the effective stress intensity range and therefore the crack growth rate.
RkJQdWJsaXNoZXIy MjM0NDE=