Curing Residual Strain Monitoring of Carbon Fiber Reinforced Polymer Composite Using Notched Long-Period Fiber Grating

Abstract

Composite materials are widely used in the aerospace industry and structural engineering owing to their advantageous mechanical properties. The curing monitoring of composite material is important to ensure the quality of the curing process, especially for the characterization of residual strains after manufacturing. In this study, we present a notched long-period fiber grating (NLPFG) with a period of 650 μm and a diameter of 66 μm that can be used in the curing monitoring of composite materials. This NLPFG was embedded into the middle layers of composite materials in order to determine the curing residual stress exhibited by the materials. The experimental results showed that the residual stress was about 107 MPa and the axial residual strain was 1490 με. Therefore, the proposed NLPFG has potential as a strain sensor for composite materials.

1. Introduction

The major advantages of carbon fiber reinforced plastics (CFRP) include low weight, high strength and stiffness, excellent fatigue strength, corrosion resistance and electrical insulating properties ¹ . During the curing process, the methods utilized to measure the curing parameters of polymeric composites include use of the differential scanning calorimeter (DSC)^2,3 and dielectric analysis ⁴ , both of which allow for the monitoring of changes in physical or chemical properties of the composites. However, these methods are complicated, expensive, and not available for in situ and real-time curing monitoring. Nonetheless, it is critical to monitor the curing process in order to improve the mechanical behavior of CFRP composites. In addition to the above methods, many optical fiber sensor techniques have been applied in monitoring the curing of CFRP composites^5,6. Owing to the small size of the sensors and their high sensitivity, these techniques have a number of obvious advantages over traditional monitoring techniques.

Long-period fiber grating (LPFG) has been used to make sensors utilized in a variety of structural, biomedical and aerospace applications, providing measurements of temperature ⁷ , strain ⁸ and concentration ⁹ , in addition to being used in smart composite^10,11 sensors. In this study, we present a sensor made from notched long-period fiber grating (NLPFG) with a period of 650 μm and a diameter of 66 μm that can be used to monitor curing strain in composite materials. The proposed NLPFG sensor is small and compatible with ordinary polymeric materials. It can be easily embedded in an internal sensing area in a composite structure without causing significant defects and provides a non-invasive means of obtaining real-time monitoring. Therefore, the NLPFG is suitable for application as an embedded sensor in CFRP materials.

In 2001, Dunkers et al. ¹² proposed monitoring the liquid composite molding (LCM) process of resin flow using LPFG, as well as monitoring the effect of curing on fiber-reinforced plastic (FRP) using LPFG. As part of their study, they embedded LPFG sensors in equal numbers of layers of glass preforms and placed them inside a plaque mold. They observed that the resonance spectrum disappeared due to the refractive index being higher than that of the LPFG covered by resin-fiber cladding and that the stress of the FRP caused the gradual disappearance of the spectrum. In 2005, Zhu et al. ¹³ proposed a new real-time optical fiber flow monitoring system based on LPFG technology. The experimental objective of their study was to monitor the effect of different refractive indices and viscosities of resin during the LCM process on the resonance spectrum. The results showed that the resonance wavelength was blue shifted as the resin viscosity and refractive index were increased. Moreover, the resonance spectrum disappeared when the refractive index of the resin exceeded that of the fiber cladding.

In 2007, Buggy et al. ¹⁴ proposed simultaneously monitored resin curing based on the use of three types of optical fiber sensor, namely, LPFG, tilted Bragg gratings, and the Fresnel refractometer. The results showed that there was high level of consistency in the refractive indices monitored by the LPFG and the Fresnel refractometer. The resulting accuracy was 6 × 10⁻³ RIU. In 2012, Marin et al. ¹⁵ monitored the fiber glass/epoxy (RTM6)-composite curing process using a sensor composed of long period grating (LPG) and fiber bragg grating (FBG). The sensor was initially tested during the curing process of an epoxy resin alone in order to assess variations in temperature and strain. After the cooling stage, the results for residual strains, which lead to curing residual stresses, could then be calculated at 2250 με.

From the above literature review, it is clear that in recent years LPFG composite curing sensors have mainly been used to measure variations in refractive indices of polymer resins. However, less effort has been invested in embedding LPFG into polymer composites to monitor the curing process in real-time. In order to fill this research gap, the present study monitored the curing process using NLPFG produced by inductively coupled plasma (ICP) etching. The NLPFG's periodic structure was generated on the surface of the optical fiber cladding, leading the transmission spectra to vary according to changes in the refractive index, temperature and strain. The curing monitoring and residual stress measurement were then assessed through changes in the NLPFG transmission spectra. The accuracy of the curing residual stress measurements thus provided by the NLPFG is presented and analyzed in this paper.

2. Theory

In this study, a periodic structure was produced on the surface of the fiber cladding layer through dry etching. When light was then transmitted within the NLPFG, a transmission spectrum was produced by the notched periodic grating. The NLPFG couples light from the fundamental core mode to the forward propagating cladding modes. The resonant wavelengths within the transmission spectrum will thus be attenuated, which may cause a dip in the transmitted spectrum. According to the coupled mode theory, the wavelength of an NLPFG under phase matching conditions can be calculated using the following formula, herein denoted as formula (1) ¹⁶ :

$$\lambda =\left(n_{eff}^{core}-n_{core}^{cladding}\right)\Lambda$$

(1)

where λ represents the resonance dip wavelength, λ is the grating period, n _eff ^core is the effective refractive index of the core layer, and n _eff ^cladding is the effective refractive index of the cladding layer. Furthermore, the transmission loss of NLPFG was defined as follows in a study by Erdogan ¹⁶ :

$$T={\cos }^2\left(k_{core-cladding}^{ac}L\right)$$

(2)

where L indicates the length of the NLPFG and k_{core-cladding} ^ac represents the ac coupling coefficient between the core mode and the cladding mode. According to formula (1), the NLPFG resonance dip wavelength is inversely proportional to the grating periods, whereas transmission loss (T) has the distribution of a cosine-squared function.

An etched NLPFG structure comprises both etched and unetched areas. Therefore, when axial tension is applied to the NLPFG, the changes in the refractive indices of the etched and unetched areas should be considered separately. Assuming that the fibers of a given NLPFG are made of homogeneous isotropic materials, with the same changes in refractive index (Δn _eff-x = Δn _eff-y = Δn _eff-r ), the following two equations can be derived by referencing the relationship between the photoelastic refractive index and strain as well as by incorporating the modalities of the etched and unetched areas ¹⁷ :

$$\begin{array}{l}\Delta n_e\left(r\right)=\frac{-p_e{[n^{eff}\left(r\right)]}^3{\epsilon }_e}{2}\\ =\frac{-p_e{[n^{eff}\left(r\right)]}^{3}F}{2E\pi r_e^2},0\le r\le r_e\end{array}$$

(3)

$$\begin{array}{l}\Delta n_u\left(r\right)=\frac{-p_e{[n^{eff}\left(r\right)]}^3{\epsilon }_u}{2}\\ =\frac{-p_e{[n^{eff}\left(r\right)]}^{3}F}{2E\pi r_e^2},0\le r\le r_u\end{array}$$

(4)

where F is the tensile force, r _e is the radius of the etched region, r _u is the radius of the unetched region and ε _e is the lateral strain of the etched region, ε _u is the lateral strain of the unetched region. Where E is the modulus of elasticity and p _e is the effective photoelastic coefficient. According to (3) and (4), given the same force, a small etching diameter causes a correspondingly large change in the refractive index.

3. Experimental Method

3.1 Processing and Fabrication

The CFRP will be closely by air pressure beyond the diaphragm and remove excess of gas in the laminate by vacuum pump simultaneously. The curing process consists of three stages (i.e., the heating, isothermal and cooling stages). The process starts with the heating stage, which in this study included the application of 6 kg/cm² of air pressure on the diaphragm and 76 cm/Hg of vacuum pressure in the mod. The heating rate was 3°C/min, with the temperature being increased from room temperature to 150°C. The second stage was the isothermal stage, in which the temperature was maintained at 150°C for 30 mins. The third stage was the cooling stage, during which the temperature was allowed to return to room temperature. The resin viscosity increased to a high level and then became stable. Moreover, the residual strain will occur at the end stage. The temperature, pressure and vacuum levels over time during the curing process are shown in Figure 1 .

Figure 1

The temperature, pressure and vacuum levels over time during the curing process

The NLPFG sensor were fabricated using ICP in order to produce etched periodic grating. The NLPFG sensor fabrication process is shown in Figure 2 . First, it should be noted that SMF (single mode fiber) 28 fibers were used to fabricate the NLPFG. To begin the process, a stripper was used to remove the buffer layer from these fibers, and then the optical fiber was pasted on the etching fixture. A buffered oxide etching (BOE) chemical was employed on the etching fibers to reduce their thicknesses to 66 mm. The fibers were then attached to a period-structured metal-plated gratings mask (amplitude mask), and ICP etching was used to produce NLPFG. The metal grating mask was designed with periods of 650 μm to achieve a wavelength position with a signal dip close to 1550 nm. Then the ICP etching was employed to fabricate the surface notched periodic structure on the etched fiber. The ICP etching used CF₄ gas as the plasma ion reactive gas. Finally, the etched device was released with acid pickling to remove the adhesive on the fiber. Two SEM photos of the grating section of the NLPFG device are shown in Figure 3 .

Figure 2

The NLPFG fabrication process

Figure 3

SEM photos of the grating section of the NLPFG (a) at 50 × magnification and (b) 500 × magnification

3.2 Experimental Setup for Composite Curing Monitoring

The 16 layer thermosetting prepregs, Carbon/Epoxy composition T300/3501, were used for lay up the composite laminate in the sequence with embedded NLPFG, as shown in Figure 4 . The layer order and location were [90^°/0^°/90^°/0^°/90^°/0^°/90^°/0^°/(NLP FG)/0^°/90^°/0^°/90^°/0^°/90^°/0^°/90^°]. The curing process was accomplished using modified diaphragm forming (MDF) with an air compressor, vacuum pump and heat chamber. The mold used in the MDF consisted of two Teflon films to prevent adhesion, as well as a polyimide-diaphragm and o-ring used for sealing and locked, as shown in Figure 4 . The up mold and down mold were placed into the thermal chamber by using an air compressor and vacuum pump for thermoforming. The light source was a superluminescent diode (SLD), and the light signal was observed by an optical spectrum analyzer (OSA). A data acquisition (DAQ) system was used to monitor the temperature during the curing process. After curing process, we calculated the residual force by comparing the transmission loss and the results of the strain correction. And then through the etching of the cross-sectional area of the fiber to calculate the residual stress and axial residual strain on the NLPFG sensor. Figure 5 shows the experimental setup for the composite curing monitoring with the NLPFG sensor.

Figure 4

Schematic diagram of NLPFG sensor embedding and mold

Figure 5

The experimental setup for the composite curing monitoring with the NLPFG sensor

4. Results and Discussion

4.1 Tensile Loading Calibration of the NLPFG Sensor

Figure 6 shows the spectra of NLPFG sensor under various tensile loadings. We can discover from the figure that when a tensile force was applied to the NLPFG, a resonant dip gradually appeared, and the depth of the resonance dip was proportional to the magnitude of the tensile force. The wavelength of the sensor was 1548.6 nm. As the tensile force was increased, the resonance dip gradually deepened until the transmission loss reached −21.43 dB under a tensile loading of 0.1813 N, before rebounding as the loading increased further, as shown in the Figure 6 inset.

Figure 6

The force calibration spectra of NLPFG sensor and in-set The NLFPG transmission loss with tensile loading variations

4.2 Monitoring the Curing Process of CFRP Composite Materials with the NLPFG Sensor

The curing process included the heating, isothermal and cooling stages. In the heating stage, the temperature was increased from 25°C to 150°C. Figure 7 shows the spectra of the NLPFG sensor during the heating stage of the curing process. At 30°C, a resonance dip was generated as a result of the pressurizing effect of the air compressor and vacuum pump, and the transmission loss was −2.83 dB. In comparison with the force calibration results, the tensile force was less than 0.1 N. Throughout the heating stage, the spectrum of the light energy varied little. The transmission loss showed a downward trend as the temperature was increased to 60°C. During the heating stage, the resonance dip wavelength exhibited a red-shift trend.

Figure 7

The spectra of the NLPFG sensor during the heating stage

Figure 8 shows the variations in the temperature–transmission loss during the heating stage of the curing process. The transmission loss revealed a diminishing trend as the temperature was increased to 60°C. This phenomenon occurred due to the fact that the increasing temperature caused the resin to melt and be coated on the NLPFG. As the temperature was increased further above 60°C, the diminishing trend of the resonance dip began to rebound because of the decreased pressure exerted on the NLPFG by the air compressor, vacuum pump and resin as the resin was transformed from a melted liquid into a gel. As the temperature was increased to 120 °C and above, the increasing slope of the transmission loss was significantly larger. This phenomenon was presumably linked to the curing cross-linking reaction. The CFRP began to solidify, causing the influence of tensile force on the NLPFG to be decreased. As the temperature was increased to 135°C and above the temperature increased to 135°C, the increasing slope took a new turn and stabilized, presumably because the most intense temperature for the crosslinking reaction was reached, after which the reaction began to ease.

Figure 8

The variations of temperature–transmission loss during the heating stage

After the heating and isothermal stages, the composite was cooled down from 150°C to room temperature. Figure 9 shows the spectra of the NLPFG sensor during the cooling process. As the temperature decreased from 150°C to 30°C, the resonance dip exhibited a shallowing trend, which was influenced by the residual stress of the cooling shrinkage. When the temperature reached 50°C, the transmission loss reached its minimum at −16.78 dB. In comparison the force calibration results, the maximum residual stress during the cooling shrinkage was 0.245–0.2548N at 30°C. Throughout the cooling process, the light energy in the spectrum was fairly stable, except for a few minor variations, and the spectrum exhibited a slow trend of shifting toward shorter wavelengths.

Figure 9

The spectra of the NLPFG sensor during the cooling stage

An observation of the variations in transmission loss are shown in Figure 10 . The NLPFG sensor's transmission loss began to decrease after 120°C; thus, the residual stress could be inferred to occur at 120°C. As the temperature cooled beyond 65°C, the decrease in the transmission loss halted, and when the NLPFG sensor had finally cooled to room temperature, its transmission loss was −15.97 dB. Regarding the results of force calibration, the internal residual force may be approximately 0.209 N. After cooling, by executing a cross-sectional area conversion of the etched fiber, it could be calculated that the residual stress of the NLPFG sensor was 107 MPa and the axial residual strain was 1490 με.

Figure 10

The variations of temperature–transmission loss during the cooling stage

5. Conclusions

In this study, we successfully used notched long-period fiber grating (NLPFG) with a period of 650 μm and a diameter of 66 μm to achieve curing monitoring of composite materials. The results showed that during the heating stage of the composite material curing, optical spectrum variations of the NLPFG were observable at approximately 60°C. This phenomenon occurred due to the fact that the increasing temperature caused the resin to melt and be coated on the NLPFG. As the temperature was increased to 120°C and above the increasing slope of transmission loss was significantly larger. This phenomenon was presumably linked to the curing cross-linking reaction. The CFRP began to solidify, causing the influence of tensile force on the NLPFG to be decreased. During the cooling process, the transmission loss of the embedded NLPFG sensor began to decrease after 120°C. Therefore, the residual stress could be inferred to occur at 120°C. Finally, based on the force calibration and the average residual stress as measured with the NLPFG, the axial residual stress was about 107 MPa and the axial residual strain was 1490 με. The proposed sensor thus provides accurate measurements for the monitoring of the curing process.