A Study on the Effect of Frequency in Delamination of CFRP

Fatigue damage is considered to be one of the most critical phenomenon, contributing to the failure of an aircraft structure. Understanding the fatigue mechanism is essential for considering the different technical conditions that influence the fatigue life of structures. In recent years, fatigue tests are accelerated in order to reduce the testing time of materials/structures being developed, which in turn can reduce the time to market. However, an ideal fatigue experiment should involve testing the structure at its service conditions. Increasing the frequency of the fatigue test can alter the fatigue mechanisms that attribute to damage initiation and propagation. One possible solution to overcome this problem is to understand the effect of frequency on fatigue. By doing so, tests can be carried out at high frequencies and influence of frequency on damage progression can be taken into account during damage prediction. The scope of this thesis is narrowed down to carbon fibre reinforced plastics and aimed at understanding the effect of frequency on damage propagation through energy principles. Fatigue can be seen as a material degradation process through which the applied work in the form of strain energy is dissipated into damage and other energy dissipation mechanisms. Experiments were performed for Mode I delamination on double cantilever beam specimens manufactured from carbon/epoxy laminate. Within a single fatigue cycle, energy dissipation (energy that is supplied during loading phase and not returned back during unloading phase) and crack growth are mutual. In other words, crack growth occurs with the consumption of dissipated energy and energy is dissipated in creating new crack surface. For simplification of calculation, energy loss and crack growth occurring in one cycle are averaged over that cycle. The averaged quantities dU/dN and da/dN are correlated with each other for different frequencies. For the calculation of strain energy loss, strain energy at maximum displacement position is calculated from the area under load-displacement plots of the fatigue cycle. For the fatigue tests carried out at a frequency of 5 Hz, assumption that the load-displacement response is linear, holds good. For higher frequencies, the response becomes non-linear due to hysteresis effect. When strain energy was calculated for all the tests with the assumption of linear P-d response, no particular trend on the effect of frequency could be observed in da/dN vs dU/dN plots. Accounting for the non linearity in load-displacement response due to hysteresis provided light on the characterization of frequency effects. The size of hysteresis loop was approximated for different frequencies based on the observations during the experiments. It was also found that, more energy was dissipated to grow a unit crack (more crack growth resistance, G∗) for a higher frequency test. Two possible mechanisms were investigated to find a suitable explanation for the observed increase in G∗ with the increase in frequency: heat dissipation and internal heat generation that would cause a rise in specimen temperature. Measurements from thermocouple and infra red camera showed that no significant temperature rise occurred in the specimen during fatigue. This confirmed that, for CFRP under mode I fatigue loading, the hysteresis energy due to high frequency fatigue testing did not heat up the specimen, but either got dissipated as heat at a rapid rate or by some other mechanism which is still not clear. When results were compared with the conceptual model formulated in this thesis, it was found that increasing the frequency increases the available energy, dU/dN and the crack growth resistance, G∗ such that crack growth rate may either increase or decrease (depending on the interaction between the two parameters). Further, a model based on quantitative measurements of heat dissipation at higher frequencies and determination of exact coordinates of the hysteresis loop evolution in load-displacement relationship of the fatigue cycles are recommended to fully understand the effect of test frequency in fatigue damage propagation.