1. Introduction

Fiber Bragg grating (FBG) sensors have been widely used in safety monitoring of public infrastructures (bridges, dams, tunnels, etc.) for their intrinsic advantages such as immunity to electromagnetic radiation, resistance to chemical corrosion, small size, high accuracy, and capability of remote operation [1,2]. However, FBG sensors have problem in magnetic field sensing because of their inherently weak magneto-optical Faraday effect [3]. In recent years, much research has been concentrated on the development of magnetic field or electric current sensors using magnetostrictive transducers or other mechanisms. Most of them were based on giant magnetostrictive material Terfenol-D and FBG [4–11].

FBG based magnetic field sensor features a great potential for large-scale multiplexing. In general design, the magnetostrictive material functions as a magnetic actuator with magnetostrictive strain as the output, while the FBG acts as a strain sensor with input from the magnetostrictive material. In this way the shift of FBG wavelength to magnetic field response is correlated. A magnetostrictive sensor using Terfenol-D and Ni₆₅Cu₃₃Fe₂ epoxy-bonded with FBG fiber was proposed for DC-current and temperature discrimination [8]. Replacing magnetostrictive alloy (MA) with magnetostrictive composite (MC), MC-FBG magnetic field sensor was proposed [9], which is based on the direct coupling of the magnetostrictive strain in an epoxy-bonded Terfenol-D particle pseudo-1–3 MC actuator with a FBG strain sensor. However, all these methods are based on bulk magnetostrictive materials, which is problematic for miniature application.

If Terfenol-D is directly coated onto a FBG fiber, the elongation of magnetostrictive film due to magnetic field would result in a shift of FBG central wavelength, and therefore the magnetic field strength can be detected. However, the associated problem is that magnetostrictive response depends on the size of materials used; response with magnetostrictive thin film could be too weak for application compared to bulk materials. Improvement of sensitivity would be a key issue for magnetostrictive thin film materials. It has been demonstrated that the sensitivity of magnetostrictive fiber sensors is increased by removing the cladding through chemical etching thus reducing the ratio between the fiber and Terfenol-D coating [10]. However, the use of chemical etching process is relatively slow and the etched fiber is rather fragile. In 2011, Smith et al. developed a sensor based on single layer magnetostrictive film of Terfenol-D sputtered onto a single mode fiber that had a femtosecond laser inscribed FBG and slot micromachined into it [11]. However, the sensitivity of that sensor is comparatively low.

In this paper, we propose a novel device for characterizing static magnetic field through machining a spiral microgroove into the FBG cladding and sputtering Terfenol-D film into the microgroove. Femtosecond laser processing was employed to ablate the microgroove. The spiral microgroove can significantly increase sensitivity through a relative increase in the ratio of magnetostrictive film to fiber cladding, and improve the bonding intensity between magnetostrictive film and the fiber cladding. Additionally, different sizes of microstructures were also investigated with the purpose of optimizing the sensing device performance.

2. Experiments

In order to enhance the sensitivity of fiber Bragg grating coated with magnetostrictive film, the fiber was pre-ablated to form a three dimensional microstructure into its cladding using femtosecond laser. Before the laser ablation, the Bragg grating was written in a standard single mode fiber (SMF28) by phase mask technique with an ultraviolet (UV) excimer laser emitting at 248 nm wavelength.

The schematic illustration of the femtosecond laser micromachining system was shown in Fig. 1. The laser system is based on a 180 fs titanium-sapphire regenerative amplifier system (IFRIT, Cyber Laser, Tokyo, Japan) operating at a fundamental wavelength of 780 nm. The maximum pulse energy was 1.08 mJ. The energy density and number of pulses at the fiber position were controlled by an optical attenuator and an electronic shutter, respectively. A laser beam with a diameter of 5 mm in Gaussian mode was focused at a depth of 60 mm from the sample surface using a microscope objective lens (SIGMA Koki, Japan). The laser can be operated with selective repetition rates of between 1 to 1000 Hz for controlling the ablation rate. The ablation start-point was positioned by high-precision XYZ-stages. All three stages were driven by linear motors, and the positioning resolution was 1μm for the X- and Y-axes and 0.2μm for the Z-axis. A high-resolution CCD camera was set on the same axis of the laser beam, which enables real-time monitoring of the laser ablation processes.

 

Fig. 1 Schematic configuration of the femtosecond laser micromachining system.

Download Full Size | PDF

Two types of 3D microstructure were designed and fabricated for sensing devices, i.e. spiral groove (simply as ‘S’) and spiral-like groove (‘SL’). The former has continuously-curved groove, while the latter has curved groove with multiple rings, as shown in Fig. 2 respectively. In the micromachining process, a specially-designed holder was employed to drive the fiber rotating around its axis. The polyimide protective layer was mechanically removed about 12 mm of length, so that the FBG cladding could be directly ablated by laser beam. During the laser machining process of the spiral groove, the fiber rotates at a constant speed; meanwhile the stage drives the fiber moving at a corresponding feed rate along the direction of the fiber axis (X-direction). The ablation depth of the spiral groove can be controlled by selecting different feed rates. For spiral-like groove, each ring-shaped groove was solely ablated, and its groove depth can be controlled by rotation speed and repeat times. The laser pulse energy was set to 0.3 mJ, repetition rate 1000 Hz. Since the lengths of inscribed FBGs are ~6 mm, the lengths of microstructures were ablated as 8mm long in direction of the fiber axis. Especially, the laser beam was horizontally given ~25μm offset from fiber axis (core), so that the deformed or chirped gating spectrum can be avoided. Finally, nine FBG samples were prepared, including five spiral microstructures (S-1…S-5), three spiral-like microstructures (SL-1…SL-3) and one non-microstructured standard FBG type (N-1), as shown in Table 1.

 

Fig. 2 Two types of 3D microstructure designed for sensing device. (a) Spiral type (S-type, with a continue curved groove). (b) Spiral-like type (SL- type, with multi-ring shaped groove).

Download Full Size | PDF

Table 1. Specifications of Tested Fibers

View Table

In order to remove the impurity resulted from laser ablation, all FBG samples were immersed in dilute HF solution for 3 min in ultrasonic cleaner. Subsequently, thin films of giant magnetostrictive material TbDyFe were sputtered on the abovementioned optical fiber samples. A BESTECH sputtering system was used to prepare these thin films. The system is specially designed for optical fiber coating. A TbDyFe target with size of 4 inch was installed to RF sources, the pre-treated optical fibers are located at a position with substrate-target distance at 150 mm. Meanwhile, two 10 × 10mm Si pieces are used as monitoring the layer in order to evaluate the deposited thickness later. Deposition power for TbDyFe targets are 150 W, which responds to a deposition rate about 0.1 nm/s. In order to assure that the comparatively uniform film was deposited, two-step process was employed. After six-hours depositing step over, all samples was turned 180 degree, then another six-hours depositing step was carried out. Finally, the depth of the deposited film was measured as ~4.6um.

The characterization of the sensors was conducted through demodulation of the FBG wavelength under the variation of magnetic field strength. Figure 3 shows the sketch of the experimental characterization system. The static magnetic field was created using a permanent magnet cluster mounted on a translation stage which was then translated in micron scale increments towards and away from the fiber. The permanent magnets used were calibrated with a magnetic flux meter that has a calibrated accuracy to within ± 0.1 mT. The N-S orientation of permanent magnets was set as parallel to the fiber axis, and the applied field strength was determined by a given separation distance between the sensor and the magnets. A static magnetic field was used because it created no temperature increase of the surroundings as was found to be the case when a DC current source was used with a coil in initial experimentation. The FBG demodulator (SM130, Micron Optics Inc., USA), also called interrogator, has four channels that can synchronously accomplish wavelength demodulation of 80 FBGs. The resolution of this demodulator is 1 pm, and its wavelength scan frequency is 1000 Hz. The measured data was recorded and sent to computer for further data analysis.

 

Fig. 3 Schematic of TbDyFe sputtered FBG sensor structure and the characterization apparatus for static magnetic field.

Download Full Size | PDF

3. Results and discussions

As an example, the spectrum of the sample S-1 was observed using an optical spectrometer. Figure 4 shows the three spectra in different statuses. It was found that, the spectrum of FBG with spiral groove has no insertion loss or evident deformation after laser ablation. However, the central wavelength was permanently shifted from 1295.446 nm to 1295.526 nm. When a magnetic field with magnitude of 100 mT was applied on the probe, the spectrum produced parallel shift ~70 pm, and the spectrum can be recovered when the field was moved away.

 

Fig. 4 The spectra of FBG in different statuses (sample S-1).

Download Full Size | PDF

Mainly, the characterization of magnetic field was carried out using demodulator (not spectrometer) under a stable room temperature. Figure 5 shows the central wavelength shift of FBGs sputtered with Terfenol-D during the increase of magnetic field up to 140 mT for different samples. Sample S-1 and SL-1 have the same groove depth of ~13.5 μm, while non-microstructured sample N-1 is compared as reference. It shows that the shifts of FBG wavelength exhibit nearly linear in the range of 0~100 mT. When the magnetic field strength increases larger than 100 mT, the increasing trend of FBGs wavelength shifts becomes weak gradually. As the magnetic field is removed, the central wavelength of the sensors can be recovered to its original state gradually in 15 seconds. That is to say, the response to magnetic field is reversible for these sensors. In the same depth and pitch of ablated microgroove, the sample with spiral microgroove is more sensitive than that of spiral-like microgroove.

 

Fig. 5 The central wavelength shifts of different FBG samples coated with Terfenol-D.

Download Full Size | PDF

It can be concluded that, the samples with microstructure are much sensitive than the non-microstructured standard FBG (N-1). In the roughly linear region, the sensitivity of spiral type S-1 is ~0.7 pm/mT, which is approximately 5 times higher than the sample N-1 (~0.14pm/mT). In this case, the ablated spiral microstructure can significantly enhance the sensitivity of the magnetic field sensor. The reason could be explained from several aspects: Firstly, a portion of material of the FBG cladding is removed by laser, the area of the intersecting surface of the fiber decreases, leading to the improvement of retractility of local fiber. Secondly, the spiral microgroove has two side surfaces and a bottom surface, the whole area of the sputtered film increases relatively, leading to enhancement of the magnetostrictive effect. Lastly, due to the wriggle of the microgroove surface, the bonding intensity between magnetostrictive film and the fiber cladding increases, the longitudinal component of the magnetostrictive strain can act more directly on FBG, resulting in more shift of FBG wavelength. For the spiral-like type sensor, its microgroove is ring-shaped, not along fiber axis, the longitudinal component of the strain resulted from applied magnetic field should be smaller than that of the spiral type. That is why the spiral-like type sensor has lower sensitivity than the spiral type sensor.

Figure 6 plots the central wavelength shift trend of FBG samples with different depths and pitch of microgroove. Apparently, the depth of microgroove has significant effect on the sensitivity of the magnetic sensors. It can be observed from Fig. 6(a) that the deeper microgroove demonstrates the higher sensitivity to the magnetic field. The main reason is that, the deeper microgroove can enhance the retractility or flexibility of fiber cladding. However, if the microgroove is too deep, the tensile strength and bending strength will be reduced, and therefore the micro-structured fiber is not mechanically robust for application. For this reason, the depth of the microgroove should be controlled under 18 μm. In Fig. 6(b), five sensor samples have different sensitivity to magnetic field. For spiral type samples, the sensor with spiral pitch of 50 μm has the strongest response to magnetic field, followed by the pitch of 30 μm and 80 μm. It indicates that, there might be an optimum combination of groove depth and pitch for spiral type sensor. As for spiral-like type sensor, the sample with pitch of 50 μm has higher sensitivity than that with pitch of 80 μm. When the pitch of spiral-like microgroove is fabricated as 30 μm, longer processing time will required, and especially the fiber is easy to break, therefore is not practical for use as sensor structure.

 

Fig. 6 The central wavelength shift trend of FBG samples with different sizes. (a) Different depths of microgroove, (b) Different pitch of microgroove.

Download Full Size | PDF

It can be concluded from Fig. 4 and Fig. 5 that, the spiral type sensor has higher sensitivity than the spiral-like type sensor. In addition, the laser ablation of the spiral microgroove spent no longer than 7 minutes, which is very efficient. Therefore, the spiral microstructure is more suitable for magnetic field sensor.

4. Conclusions

A new type magnetic field sensor based on fiber Bragg grating with spiral microstructure and Terfenol-D coating have been proposed and demonstrated. Using femtosecond laser and a specially-designed holder, the spiral microstructure was ablated into the FBG cladding, so that the sensitivity of Terfenol-D coated FBG sensor was enhanced greatly. Three types of FBG sensor were fabricated and tested, including spiral type, spiral-like type and non-microstructured standard FBG type. The spiral type sensor has the strongest response to magnetic field, followed by spiral-like type sensor, non-microstructured standard FBG type. For these sensors, the response to magnetic field is reversible. The laser ablation of the spiral microgroove is very efficient, which is quite promising for magnetic sensor application.