An Investigation on Flexural Fatigue Behavior of CFRP Quasi-Isotropic Laminates

Out-of-plane loads are a serious concern in the area of carbon fiber reinforced plastics (CFRP) that are employed in critical structural applications. Either the impact or the flexure strength of the laminates is the limiting factor in several cases. Most of the recently developed fighter aircrafts are made out of thin skinned CFRP laminates. Overall wing being in the form of cantilever beam, the bottom skin undergoes tensile stress, while the upper skin under goes the compression stress. The wings are designed with multiple spar constructions and the aerodynamic loads are transferred to the spars through wing skins. Damage to the wing skin would affect the load transfers and service life. Hence, the wing skin materials need to be assessed for static, fatigue and impact loads. Generally, a flexure test is believed to simulate near realistic loading conditions on the material of the wing skin, as it exerts, both, the compressive and the tensile stress on the specimen under test. A flexure test can either be a three-point or a four point-bend type. A 4-point bend load test, produces more consistent results than that of a 3-point bend load due to the nature of application of the load.

In this study, AS4/914 grade CFRP laminates with two different quasi-isotropic (QI) layup sequences were compared for their performance under flexure fatigue loads. The QI laminates were designated and fabricated as Laminate-1 (L1) [0/45/-45/90]_2S and Laminate-2 (L2) [0/90/14/-45]_2S, respectively. The prepreg lamina is of 0.15 mm thickness and each laminate has 16 layers. In each case, the topmost layer (16^th layer) was in compression side while the bottommost layer (1^st layer) was in tension side. These laminates were designed, such that the 0^◦ layers are placed at a similar position in both the laminate systems by changing the other layers. Standard specimens were machined from the laminates and tested for static and fatigue loading, till final failure. An articulated 4-point bend fixture was designed and employed on a computer controlled servo-hydraulic fatigue test system.  During the test, load and displacement data was monitored online along with instantaneous number of constant load amplitude (CLA) fatigue cycles. Three load levels of 90%, 80%, and 70% of the ultimate flexure strength (UFS) were chosen for assessing the flexure fatigue behavior of the laminates.  At each of the load level at least seven specimens were tested.  A few tests were also attempted under high amplitude cycles followed by low and low amplitude cycles followed by high amplitude to examine their effect on fatigue life, as compared to the fatigue life under CLA.

For the composite considered in the study, progressive failure mechanism is observed in Laminate-2 even under static load. All the fatigue failures in both the laminates initiated from the top layer (compression side) owing to stress concentration developed by the metallic roller on the composite surface. At lower load amplitudes, in laminate-1, the top layer failure is successively followed by the subsequent layers till failure of the bottom most layer by delamination.  For similar load conditions, in laminate 2, delamination initiates between layers 14 and 13 (45/-45 compression) followed by multiple layer delamination. Observation shows that in 90% UFS fatigue load case L1 has higher flexure fatigue life than that of L2. In 80% UFS load case L1 and L2 gives same flexure fatigue life; whereas in 70% UFS load case, L2 gives about three times flexure fatigue life than that of L1. In Laminate-1 high-low and low-high load sequence appears to follow the linear damage summation rule with 10-25% lesser damage life, whereas in Laminate-2, the same appears to be in good agreement.  The results obtained in the form of data plots and failure modes supported by microscopic images of the failed surfaces and cross-sections are discussed in the paper.