Engineering composites and especially long fibre carbon composites have been in high demand not only in aerospace and automotive applications, but also in high end everyday applications. In aerospace, carbon composites are used predominantly for secondary structures attached by joints or fasteners to various alloys or even different composites, and are exposed to service loads and repetitive impacting. Impact fatigue (IF) is not studied adequately for long cycles and relevant literature is investigating mainly drop weight tests and high speed projectile experiments. The main aim of this research was to investigate the behaviour long fibre CFRP'S exposed to repeated low-velocity, low energy impacts, and to observe the damage effects of this regime on the structural integrity of these materials. Two types of specimen configurations using CFRPS's were used and exposed to loading conditions relevant to the Izod impact fatigue test (IIFT), and the tensile impact fatigue test (TIFT), in order to determine the fatigue behaviour of the specimens for each of these load conditions. For the IIFT, the fatigue life was investigated using IM7/8552 unidirectional specimens and T700/LTM45 cross-ply specimens were utilised for the TIFT. The specimen thicknesses were altered in both cases and parametric studies were carried out, where it was seen that IF results in high level of scatter and the apparent decrease in life was seen at relatively modest levels of maximum force after relatively few cycles. In the case of the IIFT, a durability limit was not apparent which increases the complications when designing against IF. In the case of the TIFT the stiffness deterioration was reflected as an increase of the loading time, in the force vs time graph, over the total fatigue life span. Fatigue crack growth was investigated using fractography and X-ray micro-CT at the micro and macro level. It was seen, that IF had the potential to initiate cracks and to cause their propagation at low levels of loading. For the IIFT, a single crack was growing substantially in the fibre direction and across the sample width causing matrix cracking and probably breaking of some fibres, which acted as impact wave guides since matrix cracks were propagating initially along the length of the fibres. In the case of the TIFT multiple damage modes were presented (matrix cracks, axial splits and delaminations). Their sequence and progression was successfully v captured and contrasted against the number of impacts. Axial splits governed the damage scenario, with delaminations extending between them and the free edges. For the TIFT, IF was studied using the force-life (F-Nf) and energy-life (E-Nf) curves. The tests undertaken showed that when halving the thickness of the laminates the fatigue life presented a 10-fold decrease as well as higher scatter. Finite element modelling was undertaken to validate the experimental data of the TIFT test. Successful simulation of a single impact was carried out using a fully transient 3-D model of the actual experiment configuration which involved geometric non-linearities in addition to the multiple contact conditions. The analysis was undertaken using the Abaqus 6.11 explicit solver. Since the numerical single impact results (force vs time response) was in agreement with the experimental results, the crack modes, experimentally observed, were also incorporated in the model utilising the use of the cohesive zone elements (CZE).

The main aim of this research was to investigate the behaviour of adhesive joints exposed to repeated low-velocity impact i.e. impact fatigue (IF), and to compare this loading regime with standard fatigue (SF), i.e. non-impacting, constant amplitude, sinusoidal loading conditions. Two types of lap joint configuration using rubber toughened modified epoxy adhesives were used and exposed to various loading conditions in order to determine the fatigue behaviour of the joints for each load conditions. The fatigue life was investigated using bonded aluminium alloy (7075-T6) single lap joint (SLJ) specimens, where it was seen that IF is an extremely damaging load regime compared to SF. Different trends were visible in force-life plots for these two types of loading. In SF a gradual decrease in the fatigue life with increasing load was observed, whereas, in IF a significant decrease in life was seen at relatively modest levels of maximum force after relatively few cycles. Comparisons of the fatigue life show a considerably earlier failure in IF than in SF for comparable levels of force and energy. Additionally, it was demonstrated that the maximum force per cycle, loading time, stiffness and strength decreased as a result of damage generated in the sample during IF.

Compressors that are used in refrigerators and air conditioners usually have flapper vales made of martensitic stainless steel to control the flow of the refrigerant in the system. During service the flapper valves are affected by both bending and impact fatigue in the very high cycle fatigue (VHCF) range with billions of cycles until failure. Due to the VHCF, it is time consuming and expensive to test the performance of the flapper valves. One approach to improve the valve testing could be to combine traditional sample testing with the finite element method (FEM). In this paper, FEM was used to calculate the velocity and stress between a flapper valve and the seat during impact. Three different valve tongue shapes were investigated: a circular and two elliptically shaped tongues with a width to length ratio of 3:2 and 2:1. Furthermore, two different load cases were used to make the valve move: a backpressure case that was adapted from a compressor manufacturer and a springback case that was adapted from a flapper valve testing platform. A study of the wear mark evolution was also made on the surface of flapper valves that impacts with the seat. The valves had been in use for a different amount of cycles, were supplied by a compressor manufacturer and were made of Sandvik Hiflex steel. Stereo microscopy, scanning electron microscopy (SEM) and surface measurements were used at three different areas on the valves. It was shown with the FEM that the maximum compressive stress, at a specific point in the material, does not occur at the impact for that point. Rather, that constructive interference between stress waves in the material is the probable cause for the stress peaks that are formed. In what way the valve impacts with the seat will affect the maximum compressive stress distribution in the valve tongue. If an area close to the root of the valve impacts first with the seat, a whiplash effect will cause a higher impact velocity and impact stress in the free end of the valve. The wear mark study showed an initially high growth rate for the wear mark. However, with an increasing amount of cycles, the wear mark growth rate will decrease. Areas at the edge of the valve tongue consistently had the lowest wear mark depth, while areas close to the root and the free end of the valve had similar wear mark depth in the longest tested valve. FEM and wear mark results indicate that the impact velocity and maximum compressive stress are important factors for wear mark growth.