The degradation of onshore, reinforced-concrete wind turbine foundations is usually assessed via above-ground inspections, or through lengthy excavation campaigns that suspend wind power generation. Foundation cracks can and do occur below ground level, and while sustained measurements of crack behaviour could be used to quantify the risk of water ingress and reinforcement corrosion, these cracks have not yet been monitored during turbine operation. Here, we outline the design, fabrication and field installation of subterranean fibre-optic sensors for monitoring the opening and lateral displacements of foundation cracks during wind turbine operation. We detail methods for in situ sensor characterisation, verify sensor responses against theoretical tower strains derived from wind speed data, and then show that measured crack displacements correlate with monitored tower strains. Our results show that foundation crack opening displacements respond linearly to tower strain and do not change by more than ±5 $\mathsf{\mu }$ m. Lateral crack displacements were found to be negligible. We anticipate that the work outlined here will provide a starting point for real-time, long-term and dynamic analyses of crack displacements in future. Our findings could furthermore inform the development of cost-effective monitoring systems for ageing wind turbine foundations.

In this work we present the first large-scale demonstration of metal packaged fibre Bragg grating sensors developed to monitor prestress levels in prestressed concrete. To validate the technology, strain and temperature sensors were mounted on steel prestressing strands in concrete beams and stressed up to 60% of the ultimate tensile strength of the strand. We discuss the methods and calibration procedures used to fabricate and attach the temperature and strain sensors. The use of induction brazing for packaging the fibre Bragg gratings and welding the sensors to prestressing strands eliminates the use of epoxy, making the technique suitable for high-stress monitoring in an irradiated, harsh industrial environment. Initial results based on the first week of data after stressing the beams show the strain sensors are able to monitor prestress levels in ambient conditions.

This paper presents experimental results from fatigue and static loading tests performed on both epoxy and braze-welded FBG strain sensors. Most FBG attachment methods are relatively understudied, with epoxy the most commonly used. Long curing times and humidity sensitivity during curing render epoxy inappropriate for certain implementations. This work shows that a bespoke braze-welded attachment design is able to achieve a higher static failure limit of 22kN when compared to strain gauge epoxies, which fail at 20kN. Both methods demonstrate high fatigue life, with no significant deterioration after two million cycles. Epoxy swelling was observed when the sensors were held at a relative humidity of 96%, applying ~0.6 me of tension to the FBG, whereas a braze-weld attachment was unaffected by humidity.

Metal packaged fibre Bragg grating sensors have the potential to provide reliable measurements of temperature and strain in high stress environments for the purpose of structural health monitoring. However due to the induction brazing fabrication process a small percentage of sensors are found to have defective brazed joints. We demonstrate a defect identification procedure derived from the batch temperature calibration of fibre Bragg grating sensors for temperature and strain measurement, allowing defective sensors to be identified before installation. The procedure was demonstrated on a sample of twelve temperature sensors fabricated for a small-scale smart pre-stressing strand validation test.

In this work, we demonstrate active and passive methods for in-situ temperature calibration of fibre Bragg grating strain and temperature sensors. The method is suitable for characterizing sensors which are already attached to the steel reinforcements of civil structures. The proposed method, which involves the use of active induction heating or passive room temperature fluctuations, can be implemented using portable equipment, is time efficient, and can be used to calibrate attached sensors on-site, rather than in lab conditions. Preliminary results of the induction heating calibration show good agreement with pre-calibrated temperature sensors. In-situ calibration of fibre strain sensors, attached to a prestressing strand is also successfully carried out