ICF10B

All fatigue tests were conducted under fully reversed loading conditions. Fully reversed fatigue testing of composites is relatively uncommon due to buckling instability issues. Anti-buckling guides may be used, but problems like frictional ware and thermal influences at elevated temperatures are also formidable [2]. The current research used thick laminates with a refined dogbone specimen design (Fig. 3) and an experimental setup that served to significantly reduce specimen instabilities. The design viability was verified by comparing the compressive strength results to those obtained using the standardized Celanese fixture (ASTM Standard D 3410). Ultimate strength results showed good agreement [3]. The specimen stability and experimental set-up were further evaluated by examining the compressive behavior while monitoring the strains on opposite faces of the sample. Deviation of these two measurements is a good indicator of the most likely state of bending. This result is shown in Fig. 4 along with the extensometer strain measurement. The results indicate minimal bending until immediately prior to failure. Further, the maximum fatigue strain ranges examined do not exceed ±0.7 % which are well within the regime absent of significant bending. 5.72 REF. 1.90 +0.020 -0.000 +/-0.002 Parallel A 7.18 R gage section 1.90 ±0.020 15.2 REF 2.03 +0.000 - 0.020 All Dimension in cm Strain, m/m -0.012 -0.010 -0.008 -0.006 -0.004 -0.002 0.000 Stress, MPa -800 -600 -400 -200 0 Extensometer δrate = 8.5 x 10-3 mm/s Gage #2 x Fatigue test strain amplitudes 121 °C Gage #1 Fig. 3-Dogbone specimen geometry. Fig. 4-Static tests indicating a lack of bending at strain ranges consistent with fatigue testing. The fully reversed fatigue tests were conducted in load-control (Rσ = -1) with load levels selected to yield a specific initial strain range (±0.4, ±0.5 and ±0.6 %). This is consistent with the fact that the design criteria for the intended application is a limiting strain. Samples were subjected to a saw tooth waveform at a frequency of 2 Hz either until complete fracture or until a predetermined reduction in modulus. In the later case, cyclic fatigue was suspended and a residual compressive strength test was performed. All static residual strength tests were conducted at 121 °C with a controlled displacement rate of 8.5 x 10-3 mm/s. FATIGUE LIFE AND MODULUS DEGRADATION RESULTS The fatigue strength as a function of the number of cycles to complete failure is presented in Figs. 5 and 6. Figure 5 does this in terms of the initial applied cyclic strain range and Fig. 6 does this in terms of the applied cyclic stress range. Data are shown at the ±0.4, ±0.5 and ±0.6 %, strain range, in addition to a datum at ±0.3 % where failure did not occur and the test was terminated at approximately 6 million cycles. The data indicate that HC conditioning does not have a strong influence on the axial fatigue life though some deficit in life may be suggested at the lower load ranges. There appears to be some measure of increased scatter at the longer life/lower strain or stress ranges, as might be expected. One should note that the appearance of the data does not correspond with the 3-regime format generally observed in the results of PMC axial fatigue [4]. The first high stress/high scatter regime, traditionally associated with the probabilistic failure of the fiber, is not present in these results. Due to the fully-reversed loading conditions all of the samples ultimately failed in compression. Compressive loads drive the phenomenon of buckling induced delamination growth which control the compressive strength of the composite [5]. Note that there were clear distinctions between the 0

RkJQdWJsaXNoZXIy MjM0NDE=