PSI - Issue 5

Martin Kadlec et al. / Procedia Structural Integrity 5 (2017) 1342–1348 Petr Homola / Structural Integrity Procedia 00 (2017) 000 – 000

1347

6

4. Discussion

The location of fracture was obviously connected both with loading type and fatigue life. For the specimens with ply drop fracture (preferably occurring on milled specimens), two kinds of features typical for mode II type of loading were found – rollers and rounded cusps. Interlaminar shear loading in mode II was confirmed also by the FE model for the 0/90 layers where both upper and lower interface of the 0/90 ply was affected by shear stress concentrations in same way. According to expectations, the thinner tapering has higher stress and thus the tapering near the gauge section is always the critical one. These observations confirm that shearing of the nearest ply drops to the gauge section is one of the damage initiation mechanism for the tailored blank thermoplastic composite. The shear stress can be reduced by ending of 0/90 plies in the thicker section and +/-45 in the thinner section with higher stress; this conclusion was found also in case of epoxy laminates by Mukherjee and Varughese (2001). Therefore, it should have been ended by +/-45 ply drops that cause lower concentration due to lower longitudinal stiffness. That way, the fatigue life of double tapering alternative should have probably been improved even more. However, the mode II shear load at the ply drops was not the critical damage mechanism for all the specimens. For the specimens with out-of-ply drop fracture, scarps in a resin rich zone were found as a typical feature for the loading mode I. It proves that the opening of the laminate in the lower radius part of the tapered zone caused by tensile stress in the upper layers as was observed on the FE model. This failure mechanism driven by interlaminar normal stress was observed also by Hoa et al. (1988). It can be reduced by lowering the tapered angle. It is most probably the reason why the single taper with 6 degree angle withstood less fatigue loading than the double tapered specimens with two 3.5 degree tapers. Another reason for the different fatigue performance of these alternatives can be the fact that the number of ply drops was only 2 for double tapering and 3 for single tapering in the critical sections near the gauge. Still, the 0/90 ply drops for double tapering have high stiffness which is deteriorating due to higher shear stress concentrations and yet this alternative was better than single taper specimen with no stand-alone 0/90 ply drops. Therefore, the effect of mode I opening for larger single taper angle must have overwhelmed the effect of the ply drop shear stress and so made the single tapered alternative less fatigue durable. This was proved by the fact that out-of ply drop fractures occurred more often at lower fatigue life than ply drop fractures. Regarding the edge surface quality, fracture at a ply drop occurred more often for milled specimens and fracture out-of-tapering occurred for non-milled specimens. The rough edge surface had probably more near edge delaminations caused by the jet cutter (Shanmugam et al. (2008)). These delaminations could be the cause of mode I damage mechanism preference that was observed as the initiation mechanism for out-of-ply drop fracture mode. This mode caused generally lower fatigue lives. The damage mechanisms described above were present even for low amplitude cycling with fatigue cycles exceeding several million cycles without final fracture. This was proved by retesting of run out specimens where the cycling was stopped after several millions of cycles and they were cycled again at a higher load level. The fatigue lives of the retested specimens were up to 1000 times lower that new specimens cycled on the same load level. It demonstrates the difference between composite and metal material where metal specimens can be legally retested after 3 million cycles on a sufficiently higher load without influence on the fatigue life. The article deals with fatigue life of thermoplastic specimens with internal ply drops used for thickness reduction for more effective use of the material. Two alternatives of ply drops were evaluated and two types of test specimen extraction was used.  The quasi-static strength for single and double tapering was not statistically different. The single tapered specimens had strength of 551 MPa that is 3 % higher than for double tapered specimens.  The alternative with single tapering had in average 10 times shorter fatigue life than the alternative with double tapering. The reason for this was higher tapering angle of the former alternative. The negative effect of mode I opening for higher single taper angle overwhelmed the effect of the ply drop shear stress and so made the single tapered specimens less fatigue resistant. 5. Conclusion