The Dynamic and Flexural Behavior of Coated GFRP Rebars after Exposure to Elevated Temperatures

Abstract

The dynamic and flexural behavior of pultruded glass fiber reinforced polymer GFRP rebars were investigated after exposure to elevated temperatures ranging from room temperature to 290 °C. The rebars were cut and grouped into two sets. The first set contained the uncoated specimens, and the second set enclosed the specimens that had been coated with a gun-sprayed thin layer of silicon matrix mixed with ceramic nanoparticles as an insulation medium. All specimens’ dynamic and flexural performances were experimentally performed after heating them inside the oven for 6 h and later cooled down outside the oven at room temperature for 24 h. The dynamic results of the coated specimens showed slight changes in the damping ratio due to the effectiveness of the coating layer. In contrast, the uncoated specimens showed a gradual increase in the damping ratio ranging from 12.5% to 43.1%. Similarly, the tested flexural strength of the coated specimens provided slight changes within the elevated temperatures, while the uncoated specimens showed a gradual decrease ranging from 3.9% to 6.4%.

1. Introduction

Composite structures made from fiber-reinforced plastic FRP materials are gradually replacing conventional structures made from metals and other materials thanks to their superior and tailored mechanical characteristics, including strength, light weight, and good dynamic damping. Civil structures such as pedestrian bridges are made nowadays from FRP rebars, and they are expected to perform better as a reinforcement than steel rebars as the GFRP rebars bonds very well to the concrete, and both will resist tensile, compressive, and shear loads. However, FRPs do not perform well in applications with elevated temperatures due to the low Tg of the common commercial matrix materials. This Tg indicates the transition of the polymer from its amorphous rigid state to a more flexible state before reaching the melting temperature (Tm). When the matrix within the laminate reaches its Tg, the laminate will start losing its integrity as a rigid structure, which will eventually decrease its resistance to external loading. Thus, the thermomechanical characteristics of the FRPs are mainly governed by the thermal properties of the matrix materials because of their low and limited values of the modulus of elasticity and glass transition temperature. The effect of these interdependent thermomechanical properties has limited the usage of FRPs in applications, with an operational temperature commonly ranging from 60 to 140 °C. Hence, any elevation beyond the matrix glass transmission temperature will cause the composite structure to lose its integrity and transform from its rigid state to a more soft and rubbery state, eventually leading to failure [1,2]. Therefore, any degree of improvement in the thermal resistance of the FRPs will result in a new application that can benefit from the FRPs superior mechanical properties and light weight. Thus, this will serve well in the advanced civil and military applications where high-performance materials are demanded.

Several studies in the composite materials micro and macro scales have investigated the possibility of upgrading the thermal performance of composite materials in elevated operational or environmental temperatures. In the micro-scale coating, L. Zhang et al. coated single carbon fibers with layers of pyrolytic carbon, SiC, fluoridated hydroxyapatite, and hydroxyapatite by chemical vapor deposition (CVD) as insulation around the fibers to provide better corrosion resistance [3]. As a result, ceramic materials might be good candidates for coating FRPs because they resist both wear and high temperatures [4,5,6,7]. Furthermore, the nano-ceramic coating was also investigated as insulation for FRP from elevated temperatures and wear. A. Bu et al. coated the surface of a T300 woven fabric with a thin layer of metal/ceramic nano-coating (Ni-P/SiC) by plasma electrolytic spraying and electroless plating that provided good resistance to high-temperature oxidation [8]. In the macro scale, in the absence of insulation, B. Yu and V.K.R. Kodur tested the tensile strength of the bond of the near-surface-mounted (NSM) FRP in a range of an elevated temperature (20–200 °C) and concluded that the NSM modulus and strength were decreased by 80–70% at 200 °C [9]. Additionally, for a lower range of elevated temperature without insulation, J.P. Firmo et al. mechanically tested the shear strength of the bond between carbon fiber-reinforced polymer (CFRP) and concrete in elevated temperatures of 20 °C, 55 °C, 90 °C, and 120 °C and found decrees in the shear strength due to adhesive softening [10]. The same is observed in mechanically testing glass fiber reinforced polymer pultruded GFRP rebars under elevated temperatures in tensile, shear, and compression loading [11,12,13]. On the other hand, the insulated FRP protected FRP confined concrete structures from fire hazards and elevated temperatures [14,15,16]. K. Dong et al. studied the performance of three insulation materials with different insulating properties. All of them succeeded in delaying the adhesive failure between the CFRP and the concrete [17].

The effect of insulating FRP laminate from elevated temperatures on the mechanical performance was confirmed when Y. K. Guruprasad and A. Ramaswamy wrapped samples of concrete cylinders with CFRP and geopolymer mortar and other samples with CFRP and ceramic fiber blanket and exposed them to high temperatures (400–715 °C) [18]. Regarding the glass transition temperature improvement, M. A. Sawpan et al. obtained the glass transition temperature of an accelerated hygrothermal-aged sample of pultruded GFRP rebars for different aging intervals. They found a maximum increase of 6 °C due to moisture content [19]. Therefore, the proposed novel process of coating the outer layers of the FRPs with thin films of ceramic materials such as ZnO or Al₂O₃ in the form of powdered nanoparticles by PVD will act as an insulation medium between the composite laminate and the external heat sources. Moreover, the thin ceramic coat may increase the laminate wear resistance to harsh chemicals and may at least double its hardness. Additionally, for applications that involve fluid movement, the coat’s smooth surfaces may improve the fluidity of the moving fluids, which will lead to energy savings.

Dynamic properties also have a significant impact on the behavior of composite structures and have therefore been examined in several studies; hence, the experimental dynamic behavior of GFRP composite takes many tests according to their different types and purposes; for example, the drop hammer test was used to study the influence of varied loading rates on the ultimate compressive stress in GFRP bars. In this test, one end of the bar was kept together using artificial clay. Additionally, a dynamic force transducer was used to determine the axial force acting on the bar specimens, an accelerometer was used to determine acceleration, and a laser displacement transducer was utilized to determine the resultant deflection [20,21]. The drop hammer test revealed that most GFRP bars failed due to fiber splitting, with several specimens completely smashed. Some numerical models and investigations were performed to predict the dynamic behavior of composite rebars [22,23,24]. Marur and Kant [25] developed three revised higher-order displacement models to investigate free vibration in sandwich and composite beams. The free vibrations of thin-walled laminated composite beams were studied [26]. Hamilton’s principle was used to develop their equations of motion.

Additionally, they demonstrated how bending and torsional modes might be coupled in any laminate stack topologies. In contrast, dynamic vibration tests such as continuous monitoring of the structure’s vibration behavior of GFRP in an I-section have been used [27]. Vertical accelerations were recorded at midspan with two accelerometers attached to a data acquisition device located on the underside of the GFRP sections. Data were collected in a time of 0.01 s, which allowed frequencies up to 50 Hz to be recorded. The dynamic behavior of the UFSC panel system was investigated concerning vibrations induced by human movements such as walking and jumping. The free vibration of composite structures is widely used to characterize the dynamic properties [28,29,30,31,32].

The influence of the free vibration test on the GFRP reinforcement bars to determine the resonant frequency and damping capacity with and without ceramic coating at different elevated temperatures has not yet been investigated. Therefore, this study aims to experimentally measure the rebar’s natural frequency, damping ratio, and dynamic modulus.

2. Experimental Procedure

The coated and the uncoated rebar specimens were prepared in two sets. One set was for the 3-point flexural test, and the other was for the free vibration test. In addition to room temperature T1, both sets of specimens were conditioned in a vacuum drying oven (DZF-6050, Shanghai Zhongyu Instrument Equipment Co., Ltd., Shanghai, China) for different pre-set high ambient temperatures above the specimens’ glass transition temperature Tg (T2: 140 °C, T3: 200 °C, T4: 290 °C) for a continuous 6 h (see Figure 1). The oven was programmed to start heating from room temperature until it reached the pre-set elevated temperature and keep it with a tolerance range of ±7 °C. At the end of the heating conditioning period, the specimens were extracted from the oven and left to cool down to room temperature for at least 24 h before testing. The flexural test was performed as per ASTM A370 standard on a UTM SANS (CMT5205, Sansi Yongheng Technology (Zhejiang) Co., Ltd., Ningbo, China), 300 kN testing machine with a span length of 28 cm, as shown in Figure 2. The test was performed in a displacement-control setup with a 1 mm/min crosshead speed. The dynamic behavior was investigated and analyzed by a free vibration test machine.

3. Specimens

A pultruded, helically wrapped, and rough surface reinforcing GFRP rebars with a 10 mm diameter supplied by Henan Top Industry Technology Co., Ltd. (Jiaozuo, China) were cut by a diamond blade table saw into two sets of specimens for the 3-point flexural and the free vibration tests with lengths of 300 and 130 mm, respectively. The length of the 3-point flexural test specimens was cut to 300 mm to mitigate the shear (delamination) failure and to allow the specimens to fail due to tensile/compressive fracture failure. The rebars were made from glass fibers as a reinforcement and polyester resin matrix. The supplied rebars’ properties per the manufacturer are in

Table 1. The properties of the rebars and the coating material are as per the suppliers’ data sheets.

GFRP Rebars Properties
Diameter (mm)	10
Cross section (mm2)	73
Density (g/cm3)	2.2
Weight (g/m)	150
Ultimate tensile strength (MPa)	980
Ultimate shear strength (MPa)	>150
Modulus of elasticity (GPa)	>40
Fiber volume fraction (%)	70
Coating Material Properties
Viscosity (cP)	30
Density (g/mL)	1.36
Coating Material Mixture
p-chlorobenzotrifluoride (wt %)	50–<75
Ambient Cure Refractory Resin (wt %)	25–<50
Ceramic-Based Pigments and Additives (wt %)	25–<50
Performance Ceramic #1 (wt %)	10–<25
Carbon Black (wt %)	1–<5
Rheology Modifiers (wt %)	1–<5

. A total of 35 specimens were prepared for both tests. Three identical specimens were given for each condition of the flexural test, and two identical specimens were prepared for the free vibration test. A thin layer of a ceramic coating material C-219 supplied by NIC Industries, CERAKOTE© (White City, OR, USA, coating properties are in Table 1) was gun-sprayed on 15 specimens from both sets and cured at room temperature for at least 7 days. As per the coating material manufacturer, this thin layer will function as a thermal insulating barrier between the rebars and the ambient elevated temperature conditions up to 427 °C (800 °F). The two groups of coated and uncoated specimens were coded according to the elevated temperatures shown in

Table 2. Specimen coding.

Specimen Type	Specimen Number	Temp (°C)
Uncoated	T1R	23
	T2R	140 ± 7
	T3R	200 ± 7
	T4R	290 ± 7
Coated	T2C	140 ± 7
	T3C	200 ± 7
	T4C	290 ± 7

.

4. Free Vibration Test

The specimens’ dynamic properties were obtained by using the free vibration method. The GFRP rebars were held and maintained in a cantilever beam, and a Constant Current Line Drive (CCLD) accelerometer was used to determine the storage modulus, loss factor, and vibration signals (Type 4507-B, Bruel & Kjaer, Naeuram, Denmark). Mounting clips were used to secure the accelerometers on the specimens. The excitation was accomplished by using an impact hammer equipped with a force transducer (Type 8206, Bruel & Kjaer, Naeuram, Denmark). A pulse analyzer LAN-XI was used to acquire and analyze the vibration signal (type 3050 A-60, Bruel & Kjaer, Naeuram, Denmark). Later, a Fast Fourier Transformation (FFT) with a resolution of 1600 lines and a frequency range of 10 kHz was completed. The average number of recorded data blocks in the linear averaging mode was ten. Each test was subjected to five impact attempts to generate an average result. The post-processing data were evaluated by using the ME’Scope program (V5, 2005, Vibrant Technology, Centennial, CO, USA). The dynamic characteristics of the tested specimens were estimated. Thus, the second-order system may be solved by Equation (1). The damping ratio and loss factor for an even number of cycles may be determined using Equations (3) and (4).

$$y\left(t\right)=y_{o }{e}^{-\zeta {\omega }_{n}t}\sqrt{\frac{1}{1-{\mathsf{\zeta }}^2}}\cos \left({\omega }_n\sqrt{1-{\zeta }^{2}t-\varnothing }\right)$$

(1)

$$\delta =\frac{1}{n}\ln \frac{x_o}{x_n}$$

(2)

$$\zeta =\frac{\delta }{\sqrt{4{\pi }^2+{\delta }^2}}$$

(3)

$$\eta =2\mathsf{\zeta }\sqrt{1-{\mathsf{\zeta }}^2}$$

(4)

$${\omega }_n=\frac{\sqrt{4{\pi }^{2 }+{\delta }^2}}{T}$$

(5)

The dynamic modulus for the first natural frequency mode were calculated from the formula in Equations (5) and (6) [27].

$${\omega }_n={(1.8751)}^2\sqrt{\frac{E_d I}{\rho A L^4 }}$$

(6)

where δ is the logarithmic decrement, E_d = E′ is the storage modulus of the beam material in Pascal, I is the momentum area inertia, E″ is the loss modulus, n is the resonance frequency for mode n in radians per second, t is the time decay, L is the beam length in meters, n is the mode number, A is the cross-sectional area mm², and ρ is the beam density (kg/m³).

5. Experimental Results

Dynamic Results

The dynamic properties have been calculated using a method of free vibration, which allows obtaining the dynamic modulus experimentally. The natural frequency of the specimens assessed at room temperature T1 did not exceed 181 Hz for the specimens exposed to different temperatures without the ceramic coating. However, when the temperature increased to T2, the resonant frequency also increased to 217 Hz, which reflects the dynamic modulus results, which showed a direct relationship between the dynamic modulus and resonance frequency listed in

Table 3. The dynamic properties of the tested specimens.

Specimen	Resonant Frequency Fn (Hz)	Damping Ratio ζ%	Loss Factor η%	Dynamic Modulus E (GPa)
T1R	183.35 ± 3.41	1.03 ± 0.31	2.06	48.79 ± 2.41
T2R	217.09 ± 1.89	1.81 ± 0.62	3.62	68.60 ± 3.71
T3R	188.97 ± 2.19	2.39 ± 0.32	4.78	51.49 ± 2.92
T4R	219.02 ± 2.85	2.73 ± 0.44	5.46	69.87 ± 1.85
T2C	185.18 ± 1.97	2.55 ± 0.31	5.11	49.86 ± 2.54
T3C	188.46 ± 2.65	2.02 ± 0.23	4.05	51.49 ± 2.18
T4C	160.82 ± 2.74	2.22 ± 0.13	4.45	37.29 ± 1.77

. When the temperature increased to T3, it was noticed that there was a decrease in the frequency value from 217 Hz to 188 Hz. This implies a direct decrease in the dynamic modulus value, as shown in Figure 3. This decrease in the dynamic modulus might be due to the condition when the polymer material reaches a different transition state by exhibiting low mechanical properties. In contrast, when the temperature increased to T4, the original state of the polymer at T1 noticeably transitioned to another state with a higher stiffness than the other previously recorded specimens of T1, T2, and T3. This polymeric transition, which resulted from the increase in temperature to 290 °C, led to an increase in the dynamic modulus. However, this suggests that the mechanical and dynamic properties of the tested specimens have not improved by heating them to 290 °C, but the polymer matrix becomes more brittle.

As illustrated in Figure 4a,b, the effect of heat on the dynamic behavior has a clear impact on the damping ratio values, as shown in the time decay and the corresponding resonant frequency. In the case of the coated specimens, the coating layer was apparently effective at temperatures of T2 and T3, and the thin layer coating protected the GFRP specimens from direct exposure to the heat inside the oven. Specimen T1R performed the same as T3C, which maintained its state before and after the exposure to the elevated temperatures. From Figure 5, it has been found that the elevated temperatures had affected the dynamic damping properties; as the temperatures increase, the damping coefficients also increase proportionally, which affects the GFRP stiffness and strength. As the temperature rises to T3, the frequency values fall to 188 Hz. This resulted from the direct drop-in the dynamic modulus value. This could be justified as the polymer achieves a new transition state with low mechanical characteristics. By contrast, when the temperature rose to T4, the polymer transitioned to a stiffer state, leading to a net increase in the specimen’s Youngs modulus. This does not suggest that heating to T4 would not increase the mechanical and dynamic capabilities of the polymer resin matrix; instead, the polymer resin attained some qualities such as those of brittle materials. Consequently, the findings in Figure 3a might be interpreted.

By adding a ceramic coating, it has been observed that there are slight differences in dynamic damping performance values for specimens T2C, T3C, and T4C at 2.56, 2.03, and 2.23, respectively (shown in Figure 6).

6. Flexural Test Results

The results of the uncoated specimens showed a gradual decrease in the flexural strength P_max against the gradual increase in the exposure to the elevated temperatures (T2, T3), as can be seen in

Table 4. Flexural test experimental results.

Specimen	Pmax (N)	CV%	dmax (mm)	CV%
T1R	1809	±0.3	27.5	±1.8
T2R	1737	±4.4	25.3	±5.1
T3R	1626	±4.8	23.9	±4.2
T4R	1915	±4.1	29.0	±5.9
T2C	1738	±6.4	25.8	±8.0
T3C	1727	±2.7	25.0	±3.1
T4C	1619	±4.0	23.3	±3.8

, Figure 7a, or Figure 8a. The flexural strength between T1 and T2 was slightly decreased by 4%, and by increasing the temperature to T3, the flexural strength was reduced by 6.4 and 10.1% compared to T2 and T1, respectively. In contrast, by increasing the temperature to T4, the specimens’ flexural strength was notably increased by 5.5, 9.3, and 15.1% concerning T1, T2, and T3. Additionally, the maximum flexural deflection of the uncoated specimens was also observed to follow the obtained behavior of the flexural strength by having a slight decrease in the maximum deflection d_max between T1 and T2 by 8.0% (See Figure 7b or Figure 8b). However, the T3 specimens performed the least by failing at 23.9 mm, which is 13.1 and 5.5% less than T1 and T2, respectively. The d_max of specimen T3 showed a good improvement by deferring the flexural failure by 5.17, 12.8, and 17.7% with respect to T1, T2, and T3. Therefore, the effect of the elevated temperature on GFRP structures is evidently degrading the mechanical performance such as tensile, compressive, flexural, and shear strengths, as illustrated in this study and [10,11,12,13]. On the other hand, the coated specimens showed a more stable mechanical response after the exposure to the elevated temperatures for 6 h. The thin ceramic coating layer provided good insulation between the heat energy inside the oven and the GFRP rebars, and this agrees with the results obtained by [4,5,6,7]. However, a slight gradual decrease in flexural strength and maximum deflection was recorded without a sudden increase in the tested properties, as previously observed in uncoated specimens at T4. The flexural strength has been reduced from T1 to T2, T3, and T4 by 3.9, 4.5, and 10.5%, respectively. At T3, the ceramic coating protected the rebar specimen from losing its stiffness by failing at 5.8% more than the uncoated T2. The local failure zone at the center of the span length of all tested rebars appeared to fail as a combination of fiber fracture (crushing) and delamination in the compression side of the rebar. Initial delamination failure propagated to fiber fractures on the tensile side as the testing machine crosshead moves downward (see Figure 9). The exterior color of the coating layer remained the same for all elevated temperatures. At the same time, the uncoated specimens changed their exterior to darker colors gradually from greenish-white (original resin matrix color) to black as the exposure to elevated temperatures increased, as shown in Figure 10. Moreover, the coated T4 specimen insulted the rebars and prevented the input heat energy and then cooling down of specimens from significantly decreasing the specific volume of the matrix in its amorphous state, as in the uncoated T4 specimen. These results evidently showed the impact of increasing the input heat energy on the mechanical properties of test specimens’, which might be affected by the test parameters of heating time, coat thickness, cooling rate, matrix type, and Tg, of which all or some might contribute to the reduction of the specific volume of rebars’ matrix by decreasing the free volumes positioned inside the matrix when it is in an amorphous state before exposure to the elevated temperatures [33].

7. Conclusions

• The effect of exposing the uncoated specimens to elevated temperatures has been experimentally observed and provided two different behaviors in resisting flexural loading:
  • At the T2 and T3 temperatures, the uncoated specimens lost their stiffness due to the softness of the matrix and failed at lower loads than T1 specimen.
  • Contrariwise, the T4 specimen gained more stiffness and failed by a load higher than all other specimens, which might be correlated to the reduction of the polymer matrix-specific volume.
• Although the coating layer was too thin (one layer), the uncoated rebars performed very well in insulating the T2, T3, and T4 specimens, which allowed them to lose their strength gradually without having an unpredicted major change in the strength of the rebar, as demonstrated in the uncoated T4 specimen.
• It is recommended to coat the rebars with two or more layers from the coating material to achieve more insulation.
• Furthermore, the rebar’s strength and deflection performance could be improved by heat treating them at an elevated temperature, then cooling them down to room temperature, and later insulating them with the ceramic coating material.
• The dynamic properties of the coated samples showed an enhancement in the damping capacity values; hence, the coated sample protects the GFRP against the heat, which is reflected in the impact force and wave propagation during the free vibration impact test.