In previous work on smooth specimens tested in seawater, a high frequency +/-60 MPa flutter loading superimposed on a low frequency zero-to-tension trapezoidal waveform induced early fatigue crack initiation and reduced significantly the number of fatigue cycles for total failure [1,2]. Results in air on smooth specimens, however, showed that flutter loads with frequencies as high as 30 Hz failed to initiate fatigue cracking [1]. The micromechanics of fatigue cracking in seawater were reported previously [3] for cases where the flutter load frequencies were 5 and 30 Hz. In this work the two flutter load frequencies were selected for further testing in air and their effects on fatigue crack propagation rates (da/dN vs. AK) were investigated. In addition, the effect of (AP)n~,J(AP),~ ratios on fatigue crack propagation rates at above frequencies was investigated.Material studied was a submarine pressure hull steel designated as BIS 812 EMA. Precracked compact tension type specimens (T-L orientation) of dimensions shown in Fig. 1 were cycled using the loading profile as illustrated in Fig. 2. The ramp up/down rate in the basic low frequency trapezoidal waveform was at 1.5 kN/s. Ramp up and hold portions of this waveform were executed in 60s and the flutter load amplitude was held constant in all tests. The change in (AP)nuJ(AP)~ ratio was achieved by maintaining (AP)~u,= constant and decreasing Pt~ level (see Fig. 2).Crack extension was monitored at appropriate intervals during the test by the unloading compliance method. An ASTM type clip gauge [4] was used to measure the change in compliance which was then converted to a/W ratio using the solution suggested by Saxena and Hudak [5]. The stress intensity solution used was that given by Srawley [6]. The fatigue data reduction was carried out using the incremental polynomial method as outlined in the ASTM standard fatigue crack propagation methodologies [7].Results are shown in Figs. 3 and 4 with the AK taken as AK~ (as shown in Fig. 2). The (AP)nuJ(AP)~ ratio was 0.4 in tests whose results are shown in Fig. 3 and the initial AK~ at commencement of testing was between 22.0 MPa m 1~ to 25.5 MPa mta. Figure 3 shows the crack propagation rates at 5 and 30 Hz flutter frequencies as well as the crack propagation rate determined in previous work without superimposed flutter loading [8]. Results showed that fatigue crack propagation rates were flutter load frequency dependent and increased with Int Journ of Fracture 66 (1994)
