3.2. Fatigue Monitoring
The distributed strain measured along Fibre 1 and Fibre 2 when the structure was subjected to different loading regime is reported in Figure 7
. It can be seen that the measurement obtained with Fibre 2 is essentially a mirror image of the strain distribution measured by Fibre 1. The strain profiles measured by both fibres indicate a consistent strain measurement by the distributed optical fibre sensors. The strain measurement also shows uniform deformation along both edges of the specimen. The consistency of the measured strain along Fibre 1 and 2 confirms that an adequate bond between the fibre and host structure (composite specimen) is achieved. Furthermore, the strain measured from both fibres show that the top region of the specimen is experiencing a higher strain than the bottom region. This is attributed to the geometry of the test specimen (Figure 3
Ideally, the specimen should experience uniform strain along the entire specimen under tensile testing, but from the observation (Figure 7
), it clearly shows that the top region has greater strain than the bottom of the specimen partly due to the localised change in adherend thickness provided by the stepped adherend geometry. Additionally, the result shown in Figure 7
shows a small strain measurement being detected at the “looped” region which is in between Fibre 1 and 2. This is because the fibre section between Fibre 1 and 2 was bonded to the surface of the host structure with aim of avoiding high strain gradient. As the fibre can only measure the mechanical strain in axial direction, the fibre bonded in the loop will still experience the strain in the “loop” direction, however this is not an area of interest. The results presented showed that a reliable strain distribution can be measured.
The entire fatigue test was conducted for approximately 14 h. The strain measured at the centre point of Fibre 1 over the last 100,000 cycles (before failure occurs) is presented in Figure 8
. The strain measurement shows that around 200,000 cycles, the strain measurement detected a sudden increase in the tension-tension block loading regime (2000 µε to the next tension-tension block loading regime with a strain amplitude of 3000 µε). The INSTRON machine was programmed to stop for one minute before it moves on to the next level of tension–tension block loading cycle. This is clearly detected by the distributed optical fibre sensors. This result demonstrates the capability of distributed optical fibre sensor to provide in situ
monitoring of a structure.
The failure of the stepped lap joint composite specimen occurs after 254,362 cycles in this fatigue loading test. As highlighted by the dotted region in Figure 8
and Figure 9
, the strain gradually changes in the last 3500 cycles before failure occurs. It is important to note that the distributed optical fibre sensors was able to record the change in strain measured before failure occurs. These features can be used as an indication of damage or crack growth. These results demonstrate the ability of distributed strain measurement technique for the in-situ
monitoring of damage along the bonded step lap joints. The optical fibre survived the entire duration of this fatigue test as the failure strain of the optical fibre exceeds that of the adhesive joint. When the fibre is broken, the breakage point is detectable by using an optical backscattered reflectometry (OBR) [22
]. The location of the fracture point along the optical fibres can be used as an indication of damage and cracks along the structure.
A 3D plot presented in Figure 10
shows the distributed strain measured along Fibre 1 during the 3000 µε block loading regime prior to specimen failure. This is one of the characteristics of the distributed optical fibre sensors which can be used to monitor structural health in term of both distributed spatial and temporal information. This shows the capability of DOFS to perform a real-time and continuous monitoring of the measured structure. In Figure 10
, the results also show a significant change in strain measurement along the entire bonded Fibre 1 during the last 4000 loading cycles. The structure failed at 254,362 cycles during the 3000 µε loading regime. Figure 11
a shows the composite specimen after final failure has occurred. The bonded joint configuration showed first ply failure followed by net-tension failure as well as the adhesive failure, Figure 11
The distributed strain measurement technique is implemented to monitor bondline cracking during the fatigue test. High stress concentrations are found at the ends of the bondline overlap and hence it is expected bondline cracking will originate from these regions. A more detailed analysis of the strain measurement along Fibre 1 was looked at in the last 50,000 cycles before final failure. The difference in strain measured along Fibre 1 at the bondeline overlap was calculated using: Δε = corresponding strain measurement—reference strain measurement (at 200,000 cycles). This was plotted over the last 50,000 cycles and is shown in Figure 12
The bonded area has a total overlap length of 90 mm. From Figure 12
, 0 mm represents the starting point of the data gathered by the fibre which is located closest to the bottom edge of the overlap whilst 90 mm is the location of the fibre at the top edge of the specimen. It can be clearly seen that the strain at the bottom edge of the specimen remains relatively constant between 220,000 cycles and 250,000 cycles. However, at the other end of the overlap (90 mm) there is a gradual increase in strain. Due to the localised change in adherend thickness from the stepped geometry, the top section of the overlap expereineces a greater localised strain measurement due to only being 4 plies thick whilst the bottom end is 22 plies thick. Furthermore, the gradual increase in strain as the number of cycles increase suggest the likelihood of damage progression i.e.
, disbond in this case. This agrees with prior understanding that damage initaition occurs at the ends of the overlap due to the high stress concentration in these outer overlap regions. Overall the repeated loading and unloading of the composite joint structure promotes damage progression. Eventually in the last 4000 cycles of the fatigue loading regime, cracks are seen from either side of the bondline overlap. This promotes greater out of plane bending resulting in a significant increase in the strain measured by the optical fibre (252,000 cycles to 254,000 cycles).
The findings overall conclude that the distributed optical fibre sensors seemed to offer a considerable promise in damage assessment and monitoring of fatigue crack growth along bondlines. This distributed strain measurement technique also offers an alternative method for detecting and monitoring a structure whilst in flight as well as on the ground. The distributed optical fibre strain measurement approach has the potential for evaluating the effects of damage with interactive and predictive capabilities. Thus, effective and efficient repair techniques can be applied to the damage structures before the likelihood of catastrophic failure.