Smart materials get even smarter

Dries Van Thourhout is a Professor at Ghent University and senior researcher at IMEC, who was the co-ordinating partner in the 'Smartfiber' project – a project that intended to use existing fibre optic sensor technology to test the condition of the material. Prof Van Thourhout explained: "The goal was to make a miniature interrogator system, including the method to embed it in composite structures." This is important for composites used, for example, in wind turbines. Prof Van Thourhout said: "The people who design these structures do not know exactly when they will break, so they overdesign them. In general, therefore, they are too heavy, or perhaps windmills never run at full capacity because the operators are afraid that they will break. At the moments, there is no good way to monitor, over a long period of time, whether or not there is breakage or damage to the blade." Fibre Bragg Grating (FBG) sensors, the data collection device used in the Smartfiber project, are an established technology. FBGs are created by modifying an optical cable with UV light. When light from a broadband source is coupled into an FBG's fibre, only a narrow spectrum of frequencies of the input light is reflected: the central wavelength of this spectrum is commonly called the Bragg wavelength. The light that is not reflected by the Bragg grating further propagates through the fibre. A mechanical strain applied to the sensor will result in a change in the period of the grating and a change in the refractive index in its proximity. The resulting relative shift in the Bragg wavelength is proportional to the applied strain and thus can be used as a measure for it. However, while embedding 125µm fibre optics in composites had been done before, even something this small could have a negative impact on the material's strength and durability. To obviate this, FBGS Technologies, one of the partners in the project, developed a draw tower fibre Bragg grating (DTG) with a 'cladded' diameter of just 60µm and researchers found this to be as strong in pull tests as commercially available fibres. While sensors had previously been hand placed in composites, a robot platform was used as the basis for an automated system in the project. A dedicated optical fibre placement head was designed and built for mounting onto the robot arm, with a view to providing a complete manufacturing flow of a composite part in the future. The main part of the project involved the development of the embedded read out unit. The core of the interrogator is a photonic integrated circuit (PIC) fabricated using a silicon photonics platform. The signal transmitted by the PIC, is picked up by means of an array of photodetectors, and then the resulting electrical signal is processed electronically before being transmitted to the external read-out unit through the wireless channel. The external read-out unit, in addition, provides the wireless power supply to the embedded interrogator. The 'optical engine' block is at the heart of the interrogator. Here, the optical signal reflected by the FBG is transformed into an electrical signal that, after further processing, will allow calculation of the position of the FBG peak. The optical engine includes the pigtailed PIC, the detectors and the read out circuit (ROIC). The light source (SLED), the detectors and a set of sensors and actuators integrated on board, are controlled by a complex programmable logic device (CPLD). The interface to the outside world is provided by the wireless sub-module – a double inductive coil – which provides both data and power transmission. The embedded interrogator unit is 10cm in diameter. Prof Van Thourhout said: "It is UFO shaped because it is designed to have as little impact on the composite as possible. What is important is not the diameter, but the thickness, we had to keep it as thin as possible and the height is now 7mm." Such a size could still have an impact on the structural integrity of the composite admitted Prof Van Thourhout. "An action point would be to further reduce the height. There are some obvious ways to reduce it by about half, but if you want to go to really thin – something like 1mm – you need to look at unpackaged chip sensors. However, this move is determined by production volume. If you have sufficiently large volumes, you can make it smaller, but if it is for limited volumes, you can't afford to work with unpackaged chips. That is the trade off we are struggling with right now." Would a system like Smartfiber ultimately be adopted in all applications that require such monitoring, or would it only be used as a design tool during prototyping and development? "It all depends on the price," said Prof Van Thourhout. "At the moment, I think it will be used for designing new composite structures. In the project, we looked at tidal turbines, which are new and at the development stage. People are looking at developing new shapes of blades and new ways of monitoring these blades. So I think that, in the first place, it will be employed to help in the design procedure. But, ideally, we want it in every blade or every large composite structure." The SmartFiber project partners The sensor system was assembled by Optocap on a printed circuit board designed by Xenics. The optical subsystem consists of a silicon photonics integrated circuit developed by imec and photodiodes and read-out ICs provided by Xenics. Fraunhofer IIS was responsible for the wireless interface. It provides power to the embedded system and at the same time reads out the acquired data at high speed. After connecting the system to an optical fibre sensor chain manufactured by FBGS International, it was cast in an epoxy shape designed specifically by Ghent University to minimise the impact on the composite material. Finally, together with the attached fibre sensor chain, it was embedded in the blade of a tidal turbine by Airborne.