Generalized degradation model and bond failure analysis of pultruded basalt/carbon/glass FRP bars and profiles in concrete environments

Abstract

In this paper, the degradation mechanisms of pultruded fiber-reinforced polymer (FRP) composites with various types of fibers and polymer matrices, including basalt, carbon, and glass fibers, as well as amine-cured and anhydride-cured epoxy matrices, styrene-cured vinyl ester matrices, and unsaturated polyester matrices, are summarized under corrosive environments. Then, the damage mechanisms of the components of pultruded FRP composites are classified into three groups, including chemical etching & leaching, hydrolysis, and physical degradation. Additionally, a generally degradation model, the hydroxyl ions diffusion-based model (HIDM), is proposed and validated using extensive test data, demonstrating good accuracy and wide applicability for pultruded FRP composites with various cross-sectional shapes. The structural safety of FRP-reinforced concrete structures will be significantly weakened when the damage depth became greater than 6% diameter of FRP bars, corresponding to a strength retention of 77.4%. Furthermore, a new bond failure criterion for pultruded FRP bars used in construction, damage depth level, is proposed to evaluate the premature deterioration and functional obsolescence of FRP-reinforced concrete structures, which could provide a unique perspective and insight for structural safety assessment.

Keywords

FRP composites degradation mechanisms fibers polymer matrix degradation analysis concrete environment

Highlights

• Damage mechanisms of polymer matrices and fibers are summarized and classified.

• A physically based degradation model is proposed and validated by test data.

• The generalized degradation model is applicable to pultruded FRP bars, tubes, and sections.

• A bond failure criterion for the deterioration evaluation of pultruded FRP bars is provided.

Introduction

Fiber-reinforced polymer (FRP) composites have been successfully applied in the aviation, military, and automotive industries. In recent decades, they have gradually been adopted as reinforcement materials for concrete structures to prevent premature deterioration caused by steel corrosion (Zhao et al., 2019). Unlike steels, FRP composites are corrosion-resistant in environments rich in chloride salts, which can lead to severe depassivation and rusting of steel reinforcements embedded in concrete. Since the 1990s, there have been numerous successful applications of FRP-reinforced concrete structure in construction, including bridge decks and girders, seawalls, ports, and docks (Gooranorimi and Nanni, 2017; Li et al., 2019).

The durability of FRP reinforcements in concrete is widely concerned by engineers. To evaluate the FRP degradation in concrete, accelerated tests are widely performed by exposing them to harsh environments, such as seawater, wet-dry cycles, and simulated concrete/seawater sea-sand concrete pore solutions. Based on these accelerated tests, researchers have proposed several empirical and semi-empirical degradation models, as illustrated in Figure 1. These models provide good predictions under their respective exposure conditions. However, various types of concrete have been developed for construction by using different types of mixing water (e.g., water and seawater), cement/cementitious admixture (e.g., Potland cement, fly ash, silica fume, Ground granulated blast-furnace slag), and fine/coarse aggregates (e.g., river sands, sea sands, gravel, coral or recycled aggregates) (Dhondy et al., 2021; Teng et al., 2019; Yang et al., 2022; Zhang et al., 2022a; Zhou et al., 2021). Naturally, the microstructures and internal environments (e.g., pore structures, moisture content and alkalinity ) in these concrete vary, significantly affecting the long-term performance of FRP reinforcements embedded in these concrete (Bazli et al., 2021; Li et al., 2021). Therefore, the available models based on empirical equation or regression analysis from specific experimental data, lack the ability to offer a universal methodology for predicting the degradation of FRP composites under various service conditions. To address this issue, it is necessary to establish a physically based generalized model for evaluating the deterioration of pultruded FRP composites in various concrete.

Figure 1.

Common Degradation Models of FRP Composites (Adapted From (Wang et al., 2017b))

On the other hand, the FRP-to-concrete bond strength is fragile to the surface degradation of the pultruded FRP bars either embedded in concrete or adhered to concrete, such as bars, tubes and sections (Liu et al., 2024; Sun et al., 2021). Extensive researches have been conducted to evaluate the FRP-to-concrete bond strength (Benmokrane et al., 2020a; Liao et al., 2022; Taha and Alnahhal, 2021; Zhang et al., 2024a). These studies demonstrated that the bond strength could be affected by various parameters including the compressive and splitting tensile strength of concrete, effective bond length, FRP stiffness, surface treatment of FRP, rib spacing, and rib height (Lu et al., 2021; Shrestha et al., 2015). However, the FRP-to-concrete bond degradation are mainly determined by the durability and surface damage depth of FRP due to long-term exposure to the environment (ACI Committee 440, 2015, 2017; Allen and Atadero, 2012; CEN/TS, 19101, 2022; Liu et al., 2022; Ortiz et al., 2023). This is because the high alkalinity in concrete can cause the degradation of FRP reinforcements as the exposure period increased (Zhao et al., 2021, 2022; 2024a; 2024b). However, the coupled effects of long-term exposure to various service environments on pultruded FRP in construction are usually considered using the environmental reduction factors or conversion factors by the current codes (Benmokrane et al., 2020b; Ceroni et al., 2018; Correia et al., 2023; Myers and Viswanath, 2006; Zhang et al., 2022b). These factors are, in essence, to reserve an adequate safety redundancy for FRP-reinforced/-strengthened concrete structures during the design service life. No definite variables are available to calculate the damage as service time increases (Huang and Aboutaha, 2010). Hence, it is necessary to propose a definite time-dependent variable for the pultruded FRP bar-to-concrete bond strength degradation.

In this paper, the damage mechanisms of polymer matrices and fibers for various types of pultruded FRP composites are summarized and classified. Then, based on their damage mechanisms, a generalized degradation model is proposed and validated using extensive test data, demonstrating good accuracy and applicability for various types of pultruded FRP composites, such as bars, laminates, plates and tubes. Additionally, a new bond failure criteria is proposed and discussed, providing a unique perspective and insight to the long-term durability considerations for the current design guidelines.

Theory and methodology

Damage mechanisms of FRP in concrete environments

FRP composites typically consist of fibers, matrices, and fiber-matrix interfaces. Accordingly, the damage mechanisms of FRP composites can be classified into three types based on their chemical composition: degradation of glass, basalt, and carbon fibers; polymer matrices (i.e., polyester, epoxy, and vinyl ester); and the fiber-matrix interfaces.

The chemical constituents of glass fiber primarily include SiO₂, Al₂O₃, CaO, and MgO, with minor components such as ZrO₂, Na₂O, and K₂O, each comprising less than 1% by weight. Previous studies (Chen et al., 2006, 2007; Wang et al., 2017a, 2017b; Wu et al., 2015; Zhao et al., 2021, 2022, 2024c) have shown that glass fibers can be etched by hydroxyl ions. Free hydroxyl ions in solutions can react with the crystal Si-O-Si in glass fibers and generate silanol (-SiOH), a loose gel which facilitates water absorption and subsequent chemical attacks (Du et al., 2024; Guo et al., 2022; Zhao et al., 2025a, 2025b), as described by equations (1)–(3).

Basalt fibers have similar chemical compositions with glass fibers but with varying weight fractions and the addition of FeO. The presence of FeO in basalt fibers compromises their durability in concrete pore solutions (Kaushik and Islam, 1995; Mehta and Monteiro, 2014). Surface degradation occurs when chloride ions (Cl^-) and oxygen (O₂) arrive at the fiber surface and react with iron ions, forming rust (Kaushik and Islam, 1995; Mehta and Monteiro, 2014), as shown in equations (4)–(6). However, carbon fibers are inert in concrete environments (i.e., alkaline or salt-alkaline pore solutions) due to their stable chemical structure, consisting of stacks of turbostratic carbon layers (Peebles, 2018).

Polyester, epoxy, and vinyl ester matrices are commonly used in FRP applications. Polyester matrices can be hydrolyzed in alkaline solutions because their ester groups are vulnerable to hydroxyl ions (Chin et al., 2001; Kootsookos and Mouritz, 2004). Similarly, the ester groups in the cross-linking nodes of epoxy matrices, cured by aliphatic or aromatic anhydrides, are also susceptible to hydrolysis. Although the cross-linking nodes in styrene-cured vinyl matrices are generally inert, the vinyl ester resin itself contains ester groups that are prone to hydrolysis in alkaline environments, as described in equation (7) (Zhao et al., 2021; Zhao et al, 2024c).

The fiber-matrix interfaces of pultruded FRP composites for construction, are typically made of coupling agents (i.e., siloxane) designed to bond fibers and matrices together, allowing polymer matrices to transfer loads to fibers efficiently. The degradation mechanism of siloxane (Si-O-C) in concrete pore solutions is similar to that of glass fibers due to the presence of -Si-O-Si-O-C structures (Zhao et al, 2024c).

In summary, glass and basalt fibers primarily will degrade due to chemical etching and leaching in concrete pore solutions, while carbon fibers remain inert in these environments. The curing agents used also have significant influence on the degradation of polymer matrices. Polyester and anhydride-cured epoxy matrices are fragile to hydrolysis in alkaline environment (Sembokuya et al., 2003), whereas amine-cured epoxy matrices exhibit better corrosion resistance but can still absorb water due to the extensive hydrophilic groups in their cross-linked network, for example, the amine (-CONH) and ether (-O-) bonds (Fang and Guo, 2023; Gao et al., 2020; Tanks et al., 2022), leading to swelling-induced physical degradations (Hojo et al., 1991; Lim et al., 2019) and slow hydrolysis at high temperatures (Fang and Guo, 2023; Hojo et al., 1998). Besides, recent studies have reported that hydrophilic groups in amine-cured epoxy resins may become cleavable sites when exposed to adverse environments, such as ultraviolet radiation (Brand et al., 2020; Long et al., 2023; Ma et al., 2017). The damage mechanisms of FRP components exposed in concrete environments are summarized in Table 1.

Table 1.

Damage Mechanisms of Pultruded FRP Composites in Concrete Environments.

FRP components	Exposure conditions	Damage mechanisms	Eq. Number	Refs
Glass fiber Basalt fiber Interface (siloxane)	Alkaline solution Or Salt-alkaline solution	$S i - O (N a) + H_{2} O \to S i O H + {N a}^{+} + {O H}^{-}$	(1) (2) (3)	(Chen et al., 2006, 2007; Wang et al., 2017b; Wang et al., 2017b; Wu et al., 2015; Zhao et al., 2021, 2022; Zhao et al, 2024c)
		$S i - O - S i + {O H}^{-} \to S i O H + {S i - O}^{-}$
		$S i - O^{-} + H_{2} O \to S i O H + {O H}^{-}$
Basalt fiber	Salt solution Or Salt-alkaline solution	${F e}^{2 +} + {C l}^{-} \to {[F e C l c o m p l e x]}^{-}$	(4) (5) (6)	(Kaushik and Islam, 1995; Mehta and Monteiro, 2014)
		${[F e C l c o m p l e x]}^{-} + {O H}^{-} \to {F e (O H)}_{2} + {C l}^{-}$
		${F e (O H)}_{2} + O_{2} + H_{2} O \to {F e (O H)}_{3} \to {F e}_{2} O_{3} \cdot n H_{2} O$
Carbon fiber	Alkaline solution, salt-alkaline solution Or salt solution	No degradation		(Peebles, 2018)
Anhydride-cured epoxyStyrene-cured vinylPolyester matrix	Alkaline solutionOrSalt-alkaline solution	$R^{'} C O O R^{″} + {O H}^{-} ⇌ R^{'} C O^{-} (O H) O R^{″} ⟶ R^{'} C O O^{-} + R^{″} O H$	(7)	(Chin et al., 2001; Kootsookos and Mouritz, 2004; Sembokuya et al., 2003; Zhao et al., 2021, Zhao et al, 2024c)
Amine-cured epoxy	Alkaline solution or salt-alkaline solution	Water uptake and swelling-induced physical degradationDissociation of secondary/tertiary amine $R^{'} - N H - R^{″}$ and ether $R^{'} - O - R^{″}$ in the amine-cured epoxy		(Arias et al., 2018; Brand et al., 2020; Fang and Guo, 2023; Gao et al., 2020; Hojo et al., 1998; Long et al., 2023; Ma et al., 2017; Tanks et al., 2022)

Note: the alkaline solution in this table denotes the common concrete environments, while the salt-alkaline solution refers to the marine concrete environments, such as seawater sea-sand concrete pore solution.

Degradation classifications

The integrity of the polymer matrix is of significance for the service performance of FRP composites. Based on extensive accelerated tests (Arias et al., 2018; Brand et al., 2020; Chen et al., 2006, 2007; Chin et al., 2001; Fang and Guo, 2023; Gao et al., 2020; Hojo et al., 1998; Kaushik and Islam, 1995; Kootsookos and Mouritz, 2004; Long et al., 2023; Ma et al., 2017; Mehta and Monteiro, 2014; Peebles, 2018; Sembokuya et al., 2003; Tanks et al., 2022; Wang et al., 2017b; Wang et al., 2017a; Wu et al., 2015; Zhao et al., 2021, 2022; Zhao et al, 2024c), three corrosion types of polymer matrices are concluded: the surface reaction type, corroded-layer-forming type, and penetration type, as illustrated in Figure 2. The surface reaction occurs when the polymeric matrix comprises simple low molecules in the main chains and cross-links, both bonded by esters, allowing the corroded part to dissolve into the immersed aqueous solutions (Hojo et al., 1991). However, when the polymer skeleton and curing agents are longer and larger, the main polymer chains tangle, retarding the dissolution of decomposed parts, and thus the matrix corrosion shifts from surface reaction to corroded-layer-forming type. The penetration type is characterized by a two-stage diffusion/reaction behavior: firstly, the environmental solution penetrates the cured resin body until reaching equilibrium, subsequently causing a rapid decrease in mechanical strength. The latter two corrosion types are dominated by the diffusion process and apply to many commonly used polymer matrices, such as aromatic amine-cured epoxy resin and styrene-cured vinyl ester resins. According to our recent study (Zhao et al., 2021, 2022, 2024c), the surface reaction type is more applicable to unsaturated polyester-based and aliphatic/alicyclic anhydride-based FRP composites with smaller repeating units of polymer chains.

Figure 2.

The Corrosion Types of Polymer Matrices (Adapted From (Hojo et al., 1991)).

Unlike tensile strength, the compressive and shear properties of FRP composites are mainly determined by the matrix properties. Any changes in the matrix properties due to increased temperature or moisture absorption will be reflected in these matrix-controlled properties of the composites (Mallick, 2018).

Generalized degradation model of pultruded FRP composites in concrete environments

Since B/C/GFRP composites are susceptible to hydroxyl ions (OH^-) in concrete pore solutions, the degradation of FRP composites can be evaluated by examining the OH^- distributions in their cross-sections. Consequently, the damage depth of FRP composites can be determined by the threshold value of OH^- (Zhao et al., 2021; Zhao et al, 2024c). As verified in Table 2 and Figure 2, a generalized degradation model, the hydroxyl ions diffusion-based model (HIDM), has been proposed and validated by the authors (Zhao et al., 2021, 2024c).

Table 2.

Classifications of Damage Mechanisms of FRP Composites.

Components of FRP composites	Conditions	Degradation classifications
Glass fibers	Alkaline or salt-alkaline solution for all fibers	Etching & leaching
Basalt fibers		Etching & leaching
Carbon fibers		No degradation
Anhydride-cured epoxy	Alkaline or salt-alkaline solution for all epoxy and resin	Hydrolysis
Amine-cured epoxy		Swelling-induced physical degradation & subsequent scission of secondary/tertiary amine and ether groups
Styrene-cured vinyl ester resin		Hydrolysis
Unsaturated polyester matrix		Hydrolysis
Siloxane interface	Alkaline or salt-alkaline solution	Etching

When the concrete pore solutions diffuse into the FRP laminates or plates from one side, as illustrated in Figure 3(b), the governing equation can be simplified as 1D self-diffusion process, which follows the Fick’s second law.

\frac{\partial c (x, t)}{\partial t} = D \frac{\partial^{2} c (x, t)}{\partial x^{2}}

(8)

where c is the OH^- concentration, x is the diffusion depth, and D is the diffusion coefficient. The analytical solution is given as follows (Zhao et al., 2021):

c = \frac{c_{0}}{2} [1 - ψ (\frac{x}{2 \sqrt{D t}})]

(9)

where

ψ (\frac{x}{2 \sqrt{D t}})

is the error function, as shown in equation (10).

ψ (\frac{x}{2 \sqrt{D t}}) = \frac{2}{\sqrt{π}} \int_{0}^{\frac{x}{2 \sqrt{D t}}} e^{- y^{2}} d y

(10)

Figure 3.

Schematic of OH^- Diffusion in the Cross-Sections of FRP (a) Bar, (b) Laminate (Adapted from Zhao et al., 2021, 2024c) and (c) Tube.

However, the governing equation regarding the FRP bars or tubes can be described by equation (11), as illustrated in Figure 3(a) and 3(c).

\frac{\partial c (r, t)}{\partial t} = D (\frac{\partial^{2} c (r, t)}{\partial r^{2}} + \frac{1}{r} \frac{\partial c (r, t)}{\partial r})

(11)

where r is radial distance from the center, t is exposure time.

The OH^- distribution across the FRP bars is shown by equation (12) (Zhao et al, 2024c).

c (r, t) = c_{s} - 2 (c_{s} - c_{0}) \sum_{n = 1}^{\infty} \frac{J_{0} (β_{n} r)}{β_{n} R_{0} J_{1} (β_{n} R_{0})} e^{- D {β_{n}}^{2} t}

(12)

where

c_{s}

is OH^- concentration at the bar surface;

c_{0}

is the initial concentration;

x_{n} = β_{n} R_{0}

(n = 1, 2, 3 …) are the zeros of the equation

J_{0} (x) = 0

The zero-order and one-order Bessel functions of the first kind ( $J_{0},$ $J_{1}$ ) are shown by equations (13) and (14), respectively.

J_{0} (x) = 1 + \frac{(- 1)}{2^{2} \times {(1!)}^{2}} x^{2} + \frac{{(- 1)}^{2}}{2^{4} \times {(2!)}^{2}} x^{4} + \cdot \cdot \cdot + \frac{{(- 1)}^{i}}{2^{2 i} \times {(i!)}^{2}} x^{2 i} + \cdot \cdot \cdot

(13)

J_{1} (x) = \frac{x}{2} + \frac{{(- 1)}^{1}}{2^{3} \times 2!} x^{3} + \frac{{(- 1)}^{2}}{2^{5} \times 2! \times 3!} x^{5} + \cdot \cdot \cdot + \frac{{(- 1)}^{i}}{2^{2 i + 1} \times i! \times (i + 1)!} x^{2 i + 1} + \cdot \cdot \cdot

(14)

Equation (15) describes the relationship between the residual intact depth

R_{t}

, diffusion depth

R_{d f}

, and initial radius

R_{0}

. The initial strength

f_{0}

and residual strength

f_{t}

of FRP composites before and after exposure, as shown in Equation (16). The strength retention equals the

R_{t} = R_{0} - R_{d f}

(15)

\frac{f_{t}}{f_{0}} = \frac{S_{t}}{S_{0}}

(16)

R_{t} = R_{0} \sqrt{\frac{f_{t}}{f_{0}}}

(17)

\frac{f_{t}}{f_{0}} = \frac{R_{t}}{R_{0}}

(18)

\frac{f_{t}}{f_{0}} = \frac{{(R_{o u} - R_{o d})}^{2} - {(R_{i n} + R_{i d})}^{2}}{{(R_{o u} - R_{i n})}^{2}}

(19)

where

S_{0}

and

S_{t}

are the initial cross-sectional area and the residual intact area ( e.g., the area marked in blue in Figure 3(a) and 3(c)) of FRP composites before and after exposure, respectively.

The strength retention of FRP bars after exposure can be calculated using equation (17), while equation (18) can be used to calculate the strength retention for FRP laminates, as illustrated in Figure 3(b). Equation (19) is applicable to the strength degradation evaluation of FRP tubes under both outer and inner exposure to solutions, as shown in Figure 3(c). The long-term durability of other FRP composites can be assessed using equation (16) based on their cross-sectional types, such as channel and H-sections.

According to our recent research (Zhao et al., 2021, 2022; Zhao et al, 2024c; Zhao et al, 2024c), the diffusion depth $R_{d f}$ at the concentration $c (R_{d f}, t)$ of 10^-7 mol/L (pH = 7) can be assumed as the maximum damage depth, and they agreed well with the experimental data. Besides, both the residual tensile strength and interlaminar shear strength (ILSS) of FRP bars which are dependent on their cross-sectional areas, conform to the maximum cross-sectional stress criteria (Zhao et al., 2021, 2022, 2024c). This is because the OH^- penetration will reduce the effective bar diameter and the cross-sectional areas of bars (also termed as residual cross-sectional area), thus decreasing the strength capacity of FRP bars.

Besides, the diffusion coefficient D at different temperatures can be calculated using equation (20) (Antoon and Koenig, 1980).

D = D_{0} \exp (- E_{d f} / (R T))

(20)

where

D_{0}

is a constant,

E_{d f}

is the activation energy for diffusion, T is the Kelvin temperature, and R is the universal gas constant.

The $D_{0}$ and $E_{d f}$ can be determined through the experimental data at two elevated temperatures, then diffusion coefficients $D$ of FRP composites at other temperatures under concrete pore solutions can be calculated according to equation (20).

Suggested bond failure criterion for FRP composites embedded in concrete due to corrosion

The initial properties of FRP composites (i.e., the guaranteed tensile strength) usually do not consider the long-term exposure to the environments. In general, concrete structures are designed based on the limit state principles (i.e., ultimate limit state and serviceability limit state) (ACI Committee 440, 2015, 2017). However, the environmental conditions uniquely affect the long-term performance of encased FRP composites in both FRP-reinforced and FRP-strengthened concrete structures. Consequently, decades of exposure to the service environments might change the dominate limit state of concrete members reinforced with FRP bars from ultimate limit state to serviceability limit criteria, especially for those FRP reinforcements that exhibit low stiffness, such as pultruded GFRP and BFRP bars. Regarding the durability considerations for FRP reinforcements, the design strength of FRP bars is used by multiplying an environmental reduction factor, ranging from 0.5 to 1.0 (ACI Committee 440, 2015, 2017). Additionally, current codes adopt the reduced resistance capacity of FRP bar-concrete members by dividing various partial factors γ ranging from 0.6 to 1.0 (CEN/TS, 19101, 2022), thus reserving redundant safety for concrete structures. However, the current guides do not present how these factors adopted in design could extend the long-term durability of FRP-concrete structures. Besides, there are no quantitative criteria for the long-term degradation of pultruded FRP bars embedded in concrete. The environmental reduction factor C_E, partial factor γ and conversion factor η_c in the available design guides are listed in Tables 3 –5.

Table 3.

Environmental Reduction Factors for Various Fibers, FRP Systems and Exposure Conditions.

Exposure conditions	Fiber type	Environmental reduction factorC_E	Reference
Concrete not exposed to earth and weather	Carbon	1.0	(ACI Committee 440, 2015)
	Glass	0.8
	Aramid	0.9
Concrete exposed to earth and weather	Carbon	0.9	(ACI Committee 440, 2015)
	Glass	0.7
	Aramid	0.8
Interior exposure	Carbon	0.95	(ACI Committee 440, 2017)
	Glass	0.75
	Aramid	0.85
Exterior exposure(Bridges, piers, unenclosed parking garages)	Carbon	0.85	(ACI Committee 440, 2017)
	Glass	0.65
	Aramid	0.75
Aggressive environment(Chemical plants, and wastewater treatment plants)	Carbon	0.85	(ACI Committee 440, 2017)
	Glass	0.50
	Aramid	0.70
Concrete both exposed and not exposed to earth or weather	Glass	0.85	(ACI Committee 440, 2023)

Table 4.

Conversion Factors η_cm for Unprotected FRP Composite Materials and Epoxy Adhesives.

Exposure classes	Conversion factor η_cm	Influence of moisture	Reference
Ⅰ	1.00	Indoor exposure with service temperature according to 1.1 (4)	(CEN/TS, 19101, 2022)
Ⅱ	0.85	Outdoor exposure with service temperature according to 1.1 (4), without (i) continuous exposure to water, (ii) permanent immersion in water, (iii) permanent exposure to a relative humidity higher than 80%, (iv) combined UV-radiation and frequent freeze-thaw cycles	(CEN/TS, 19101, 2022)
Ⅲ	0.60	Continuous exposure to water (or seawater), or permanent immersion in water (or seawater), or permanent exposure to a relative humidity higher than 80% (material temperature up to 25°C)	(CEN/TS, 19101, 2022)

Note: The above conversion factors are applicable to composite materials with glass, carbon or basalt fibers and thermoset polymer matrix of either unsaturated polyester, vinylester or epoxy, and for epoxy adhesives.

Table 5.

Conversion Factor for Temperature η_ct for FRP Composite Materials.

Properties of composite materials	Conversion factor for temperature η_ct	Reference
For fiber-dominated properties	$η_{c t} = \min {1.0 - 0.25 \cdot \frac{T_{s} - 20}{T_{g} - 20}; 1.0}$	(CEN/TS, 19101, 2022)
For matrix-dominated properties	$η_{c t} = \min {1.0 - 0.80 \cdot \frac{T_{s} - 20}{T_{g} - 20}; 1.0}$	(CEN/TS, 19101, 2022)

Note: T_s is the maximum material temperature in service conditions (in °C); T_g is the glass transition temperature (in °C). And the conversion factor, η_{c =} η_ct•η_cm.

To address this ambiguity, it is necessary to establish a definite FRP-to-concrete bond failure criterion to evaluate the service performance of FRP bar-reinforced concrete structures from the perspective of the damage depth of FRP bars. As widely known, the serviceability conditions of FRP bar-reinforced concrete structures depend on the cooperative work between FRP bars and concrete. Hence, it is crucial to maintain good bond strength at the design level to avoid premature failure during the intended service life. However, long-term exposure to concrete environment might degrade the encased FRP bars, and the surface deterioration and an increasing damage depth of FRP bars can be expected during service. Subsequently, the bond strength between concrete and FRP bars/plates will be weakened. When the damage depth increased to a threshold value, B/C/GFRP bars will lose majority of their bond capacity to reinforce or strength concrete. Based on extensive data from published literature, Zhang et al. (Zhang et al., 2024b; 2024c) studied various surface types of FRP bars, including the helically wrapped FRP bars (i.e., Figure 4(a) and 4(b)) and deformed FRP bars (i.e., Figure 4(c) and 4(d)) with different rib height and rib spacing, and concluded that optimal FRP bar-to-concrete bond strength could be achieved when the FRP bars with a rib height of around 6% bar diameter and a rib-spacing-to diameter of 1.0, were adopted, as illustrated in Figure 4.

Figure 4.

Surface Types of FRP bars and FRP-to-Concrete Bond Degradation. Note: D_bar is the Diameter of FRP Bars, D_s is the Rib Space; r_w and r_h Denote the Rib Width and Rib Height of FRP Bars, Respectively. r_d Denotes the Critical Damage Depth.

However, when the ribs and surfaces of FRP bars were corroded, the interfacial frictions and mechanical interlock between concrete and FRP bars will disappear. Subsequently, the interfacial bond strength between FRP bars and concrete cannot be guaranteed, risking the serviceability limit state of FRP bar-concrete structures, as shown in Figure 4. Herein, the damage depth of 6% bar diameter, corresponding to a strength retention of 77.4%, can be assumed as the threshold value. Therefore, the damage depth of FRP bars embedded in concrete should be less than 6% bar diameter to avoid bond failure. To make this criteria more universal for FRP bars with various diameters, the damage depth level of FRP bars was proposed to evaluate the service performance conditions of FRP bars embedded in concrete during service. The damage depth level was defined as the ratio of the damage depth to the FRP bar diameter (termed as D_bar).

Validation and discussion

To validate the applicability of HIDM, the experimental data from various types of FRP composites in published literature, including the B/C/GFRP bars, laminates and tubes, were used to compare with the predicted strengths (Bazli et al., 2020a; Wang et al., 2017b, 2020; Wu et al., 2015). The diffusion coefficients used for the following validation are summarized in the appendix Table A1.

Pultruded BFRP bars

BFRP bars exposed in common concrete pore solutions

The experimental results from Wu et al. (Wu et al., 2015) were used to validate the effectiveness of HIDM for BFRP bars exposed to common concrete pore solutions. The initial pH value of the concrete pore solution was approximately 13.0. The OH^- distributions along the radial depth are illustrated in Figure 5(a). It was observed that OH^- concentration decreased with increasing depth at temperature of 25°C, 40°C, and 55°C. Furthermore, higher temperatures significantly accelerated the diffusion process of OH^-, thereby speeding up the degradations of BFRP bars. The predicted tensile strengths agreed well with the experimental results, with a relative error (RE) ranging from 0.4% to 3.7%, as shown in Figure 5(b). The maximum RE between experimental values and predictions was 3.7%, demonstrating the accuracy of the HIDM.

Figure 5.

Predicted Results of (a) OH^- Concentration and (b) Tensile Strength of BFRP Bars in Concrete Pore Solution. Note: BT25D21 Represents the Conditioned BFRP Bars With an Exposure Period of 21 Days at 25°C. B Denotes the Basalt Fiber, T Denotes Temperature, D Denotes the Exposure Days.

BFRP bars exposed in SWSSC pore solutions

Comparisons were conducted using experimental results from Wang et al. (Wang et al., 2017b) in both normal concrete (NC) and high-performance (HP) seawater sea-sand concrete (SWSSC) pore solutions. In this experiment, 28 accelerated conditions were employed, including four temperatures (32°C, 40°C, 48°C, 55°C) and four exposure periods (21, 42, 63, 84 days) for the normal SWSSC (NC) environment, and three temperatures (25°C, 40°C, 55°C) and four exposure periods (21, 42, 63, 84 days) for the high-performance SWSSC (HP) environment. The pH values of the NC and HP pore solutions were 13.4 and 12.7, respectively. The OH^- distributions along the radial direction were obtained using HIDM, as shown in Figure 6(a) and 7(a). Subsequently, the tensile strength of BFRP bars under different conditions were calculated, as depicted in Figure 6(b) and 7(b).

Figure 6.

Predicted Results of (a) OH^- Concentration and (b) Tensile Strength of BFRP Bars in NC SWSSC Pore Solution. Note: BT32D21 Represents the Conditioned BFRP Bars with sn Exposure Period of 21 Days at 32°C. B Denotes the Basalt Fiber, T Denotes Temperature, D Denotes the Exposure Days. NC Denotes the Normal Concrete Pore Solution.

Figure 7.

Predicted Results of (a) OH^- Concentration and (b) Tensile Strength of BFRP Bars in HP SWSSC Pore Solution. Note: BT55D84 Represents the Conditioned BFRP Bars with an Exposure Period of 84 Days at 55°C. B Denotes the Basalt Fiber, T Denotes Temperature, D Denotes the Exposure Days. HP Denotes the High-Performance Concrete Pore Solution.

Compared to the experimental data under the HP SWSSC pore solutions, the RE of predicted tensile strength varied between 1.1 - 3.6%, 4.2% - 5.4%, and 2.4% - 5.1% for the BFRP bars at 32°C, 40°C, and 55°C, respectively. These predictions were in good accordance with the test results.

For the BFRP bars under the NC SWSSC pore solutions, the predictions at 32°C and 40°C were accurate compared to experimental results, as shown in Figure 6(b). However, there were significant difference between the predicted values and test data at T48D21, T48D42, and T55D21 conditions. This discrepancy can be attributed to the large dispersion at high temperatures, which lead to greater damage depths and smaller intact core cross-sectional areas. Subsequently, OH^- ions can penetrate the residual intact cross-section of FRP bars along the defected regions. In general, the HIDM provides an acceptable evaluation of the long-term mechanical strength of BFRP bars in alkaline or salt-alkaline solutions.

Pultruded CFRP bars

The shear strength degradation of pultruded CFRP bars under normal concrete (NC) and high-performance (HP) seawater sea-sand concrete (SWSSC) pore solutions (Wang et al., 2017a) was adopted to verify the accuracy of HIDM. The predicted shear strength was compared with experimental data across 24 exposure conditions, comprising combinations of three temperatures (25°C, 40°C, 55°C) and four exposure durations (21, 42, 63, 84 days) in both NC and HP pore solutions, with pH levels of 13.4 and 12.7, respectively.

The OH^- concentration distributions in CFRP bars after exposure are depicted in Figure 8(a) and 9(a). The results show that increased temperatures accelerated the diffusion processes, significantly increasing the damage depth of the CFRP bars. Consequently, shear strength decreased with increasing temperature and exposure time in both NC and HP SWSSC pore solutions, as shown in Figure 8(b) and 9(b).

Figure 8.

Predicted Results of (a) OH^- Concentration and (b) Tensile Strength of CFRP Bars in NC SWSSC Pore Solution. Note: CT25D21 Represents the Conditioned CFRP Bars with an Exposure Period of 21 Days at 25°C. C Denotes the Carbon Fiber, T Denotes Temperature, D Denotes the Exposure Days. NC Denotes the Normal Concrete Pore Solution.

Figure 9.

Predicted Results of (a) OH^- Concentration and (b) Tensile Strength of CFRP Bars in HP SWSSC Pore Solution. Note: CT55D84 Represents the Conditioned CFRP Bars with an Exposure Period of 84 days at 55°C. C Denotes the Carbon Fiber, T Denotes Temperature, D Denotes the Exposure Days. HP Denotes the High-Performance Concrete Pore Solution.

The shear values of CFRP bars predicted by HIDM closely matched the experimental results when the CFRP bars retained high strength (greater than 75%). For instance, the maximum RE between predictions and experimental data was only 4.8% when the retentions were greater than 75% in both NC and HP SWSSC solutions, as indicated in Figure 8(b) and 9(b). The test results for conditions T55D42 and T55D63 were excluded due to inconsistencies likely caused by low manufacturing quality. However, the REs increased when the CFRP bars lost most of their strength (e.g., approximately 50% strength loss). Prediction accuracy becomes more influenced by manufacture defects as the intact area of CFRP bars diminishes. Normal diffusion paths may be altered due to voids, holes, and regional defects in the cross-section, reducing the effective radius of bars and the intact thickness of FRP laminates or tubes as temperature and exposure time increase. Additionally, FRP bars, laminates, and tubes lose their service functions when the damage depth became large, significantly degrading bond strength with adjacent concrete. Therefore, predicting the mechanical strength of FRP composites loses engineering significance under severe degradation conditions.

Pultruded GFRP bars

The shear strength of pultruded GFRP bars under normal concrete (NC) and high-performance (HP) seawater sea-sand concrete (SWSSC) environments (Wang et al., 2017a) was evaluated to verify the prediction accuracy of HIDM. As mentioned in the two cases above, the accelerated tests involved three temperatures (25°C, 40°C, 55°C) and four exposure durations (21, 42, 63, and 84 days) in the NC and HP SWSSC pore solutions.

The predictions of OH^- distributions and the resultant shear strength are illustrated in Figure 10(a) and 11(a). The predicted shear strength of GFRP bars in HP SWSSC pore solutions was notably accurate, as shown in Figure 11(b). All predictions had a maximum RE of 4.6%, except for the condition T40D21, where the RE was 7%. The RE between the experimental and predicted shear strength in the NC SWSSC environment varied from 1.1% to 8.8% when the retentions remained higher than 75%. Despite the wide dispersion of experimental data due to manufacturing and testing variability, the predicted results provided by HIDM are still acceptable.

Figure 10.

Predicted Results of (a) OH^- Concentration and (b) Tensile Strength of GFRP Bars in NC SWSSC Pore Solution. Note: GT25D21 Represents the Conditioned GFRP Bars with an Exposure Period of 21 Days at 25°C. G Denotes the Glass Fiber, T Denotes Temperature, D Denotes the Exposure Days. NC Denotes the Normal Concrete Pore Solution.

Figure 11.

Predicted Results of (a) OH^- Concentration and (b) Tensile Strength of GFRP Bars in HP SWSSC Pore Solution. Note: GT25D21 Represents the Conditioned GFRP Bars with an Exposure Period of 21 days at 25°C. G Denotes the Glass Fiber, T Denotes Temperature, D Denotes the Exposure Days. HP Denotes the High-Performance Concrete Pore Solution.

Pultruded BFRP laminates

The tensile strength of pultruded BFRP laminates in alkaline solutions (Wang et al., 2020) was used to identify the applicability of HIDM for FRP composites with various cross-sectional shapes. The BFRP laminates were immersed in an alkaline solution with a pH of 13.0 at 60°C for 7, 14 30, 90, and 180 days.

Considering the 1D diffusion process in the cross section of FRP laminates, the damage depth was calculated using equation (9) and (10). The predicted strength was then compared to the experimental results, as illustrated in Figure 12. The predictions closely matched the experimental results when the tensile strength retentions were greater than 60%, with a maximum RE of 4.8%. However, the prediction accuracy decreased when the tensile strength retentions decreased sharply below 60%.

Figure 12.

Predicted Results of BFRP Laminates in Concrete Pore Solution. Note: BT60 Represents the Conditioned BFRP Laminate Under Exposure at 60°C.

Pultruded GFRP tubes

The tensile strength data of pultruded GFRP tubes with a thickness of 8 mm from (Bazli et al., 2020a) were selected to compare with the predictions from HIDM. The GFRP tubes were exposed to the SWSSC pore solutions with the pH values of 13.4 at 25°C and 60°C for 30, 90, and 180 days (termed as T25D30, T25D90 and T25D180; T60D30, T60D90, and T60D180), respectively. It is important to note that the diffusion of OH^- ions developed from both the outer and inner surfaces of the GFRP tubes. Therefore, the damage depths on both surfaces were calculated. The OH^- distributions and predicted tensile strengths are illustrated in Figure 13(a). Where compared to the test data, the predicted tensile strength values at 25°C and 60°C were acceptable, especially when the strength retention were higher than 70%, with a maximum RE of 4.8%, as shown in Figure 13(b).

Figure 13.

Predicted Results of GFRP Tubes in Seawater Sea-Sand Concrete Pore Solution. Note: GT25D30 Represents the Conditioned GFRP Tubes with an Exposure Period of 30 Days at 25°C.

I-shaped and U-shaped pultruded GFRP profiles

I-shaped and U-shaped pultruded GFRP profiles were used to verify the applicability of HIDM for FRP profiles with complex cross-sectional types. Generally, FRP composites with intricate cross-sections can be considered as combinations of FRP laminates/plates, circular shapes, and tubes. For instance, the I-shaped and U-shaped pultruded GFRP profiles illustrated in Figure 14 consist of one web plate and two flange plates. When FRP profiles were immersed into SWSSC pore solution, the OH^- distributions in each part (i.e., web and flange laminates/plates) of the I-shaped and U-shaped FRP composites can be calculated using equation (9). Subsequently, the diffusion depth (R_df) can be determined according to equation (15) and (18). Here, due to the negligible effects of OH^- variations near the plate edges on the total corroded areas, 1D diffusion was used to calculate the diffusion depth.

Figure 14.

FRP Profiles with Cross-Sections of (a) I-Shape and (b) U-Shape; (c) Three-Point Bending Test.

As illustrated in Figure 14(c), three-point bending tests can be used to evaluate the bending strength degradation of FRP profiles with complex cross-sections before and after exposure. The ultimate bending capacity of these profiles decreased due to the reduced cross-sectional area, as illustrated in Figure 14(a) and (b). Based on the maximum stress criteria, the ultimate stress in both intact and exposed FRP profiles under bending remained unchanged. Therefore, we have

\frac{P_{t}}{P_{0}} = \frac{I_{t}}{I_{c}} \cdot \frac{y_{0}}{y_{t}}

(21)

where

P_{t}

and

P_{0}

denote the ultimate force applied at mid-span (see Figure 14(c)) for exposed and reference FRP profiles, respectively, under bending tests after a given exposure time t.

I_{t}

is the effective moment of inertia of the exposed FRP beam’s cross-section, while

I_{c}

is the moment of inertia of the reference FRP beam’s cross-section.

y_{t}

and

y_{0}

are the maximum distances from the edge to the neutral axis for the exposed and reference FRP beams, respectively.

The GFRP laminate, along with I-shaped and U-shaped (channel) vinyl-based GFRP profiles with different cross-sections, were immersed in SWSSC pore solution for 90 days (Bazli et al., 2020b). After exposure, three-point bending tests were performed on these specimens, and the results were compared with control GFRP specimens, as illustrated in Figure 14. It is noteworthy that all profiles used the same fibers, matrices, mixture proportions, and manufacturing processes. The experimental results (Bazli et al., 2020b) were used to validate the applicability of the HIDM model for GFRP profiles with various cross-sectional configurations. The diffusion depth of the GFRP laminate after 90 days of exposure was calculated using equation (18). Since diffusion occurs from both sides of the GFRP laminate and the diffusion is negligible compared to the laminate thickness, the total diffusion depth R_df can be treated as twice the one-side diffusion depth of OH^- ions.

The prediction results of the I-shaped and U-shaped GFRP profiles using HIDM agreed well with experimental data, as shown in Figure 15 and Table 6. For example, after 90 days of exposure to SWSSC pore solution, the predicted bending strength retentions of the I-shaped specimen I1 and U-shaped specimen U2 against their neutral axis were 83.4% and 75.9%, 87% and 84.6%, respectively. The maximum RE was less than 6.5%. The difference between the predicted and experimental bending strength ranged from 1.8% to 8.0%. In the case of specimen I2, it can be inferred that more initial defects were present, as its experimental bending strength retention against x-axis was significantly lower than that of the other samples after the same exposure period. In summary, the mechanical strength of pultruded GFRP profiles with complex cross-sections after exposure can be effectively evaluated using the HIDM model.

Figure 15.

Validations of I-Shaped and U-Shaped GFRP Profiles Using HIDM.

Table 6.

Comparative Results of I-Shaped and U-Shaped GFRP Profiles After 90 Days of Exposure.

Cross-section type	Specimen	Cross-sectional size			I_c (mm⁴)	y₀ (mm)	Experimental data			I_t (mm⁴)	Predicted values			RE (%)
Cross-section type	Specimen	H (mm)	B (mm)	T (mm)	I_c (mm⁴)	y₀ (mm)	Max load P₀ (N)	CV (%)	Retention (%)	I_t (mm⁴)	y_t (mm)	Max load P_t (N)	Retention (%)	RE (%)
I-shape (I1)	Control	25.5	15.0	4.0	15814	12.8	6292	0.8	100
Major axis (I_x)	Exposed	24.8	14.3	3.3			5012	3.7	79.6	12843	12.4	5245	83.4	3.8
I-shape (I1)	Control	25.5	15.0	4.0	2343	7.5	3203	1.0	100
Minor axis (I_y)	Exposed	24.8	14.3	3.3			2565	2.9	80.1	1701	7.2	2432	75.9	4.2
I-shape (I2)	Control	38.3	15.0	4.0	44727	19.2	10793	2.2	100
Major axis (I_x)	Exposed	37.6	14.3	3.3			7578	3.6	70.2	36572	18.8	8979	83.2	13
I-shape (I2)	Control	38.3	15	4	2412	7.5	6577	1.26	100
Minor axis (I_y)	Exposed	37.6	14.3	3.3			4853	5.2	73.7	1741	7.15	4966	75.5	1.8
U-shape (U1)	Control	50	30	3	120836	25	4320	2.25	100
Major axis (I_x)	Exposed	49.3	29.3	2.3			3727	3.12	86.3	93428	24.67	3385	78.3	8.0
U-shape (U1)	Control	50	30	3	27478	10.36	7250	0.9	100
Minor axis (I_y)	Exposed	49.3	29.3	2.3			5131	6.5	70.7	20753	10.25	5531	76.3	5.6
U-shape (U2)	Control	50	30	5	179167	25	7659	1.89	100
Major axis (I_x)	Exposed	49.3	29.3	4.3			7161	4.4	93.5	153791	24.67	6662	87.0	6.5
U-shape (U2)	Control	50	30	5	41667	10	11002	1.1	100
Minor axis (I_y)	Exposed	49.3	29.3	4.3			8658	2.7	78.6	34873	9.89	9307	84.6	6.0

Note: The italic values are the predicted values from the HIDM.

Conclusion remarks

In this paper, the degradation mechanisms of pultruded FRP composites with various types of fibers and matrices under concrete environments are summarized. The fibers include basalt, carbon, and glass fibers, while the polymer matrices comprise the amine-cured and anhydride-cured epoxy, vinyl ester, and unsaturated polyester. The damage mechanisms of the FRP constitutions are identified and classified into three groups. Based on their damage mechanisms, a physically based generalized degradation model, the hydroxyl ions diffusion-based model (HIDM) proposed by the authors, is validated using the available test data from published literature. The HIDM demonstrates good accuracy when the FRP composites retain adequate strength to reinforce concrete, specifically with a strength retention of greater than 70%. Besides, the damage depth level is proposed for the bond failure analysis of FRP bar-concrete structures, providing a quantitative parameter and unique perspective to the current codes. The following conclusions can be drawn:

• Both basalt and glass fibers can be etched and leached under concrete environments (i.e., alkaline or salt-alkaline pore solutions), whereas carbon fibers are inert to these corrosive environments. The degradation of the interface between fibers and matrix is similar to the etching of glass fibers.

• Unsaturated polyester, styrene-cured vinyl ester, and anhydride-cured epoxy matrices can be significantly damaged by hydrolysis in alkaline environments. The degradation of amine-cured epoxy matrices under alkaline solutions usually originates from water uptake and the resultant swelling, along with the dissociation of secondary/tertiary amines and ethers in the amine-cured epoxy.

• The degradation of pultruded FRP composites under alkaline solutions can be predicted by the generalized degradation model, HIDM, regardless of the shapes of FRP composites, such as bars, tubes, and sections. The generalized degradation model is proposed based on exposure to pore solutions, and its application in relative humidity environment are still needed using a shift equation, that is, liquid-gas-state shift theory (Zhao et al, 2024c), which will be present in our following research.

• The structural safety of FRP-reinforced concrete structures will be significantly weakened when the damage depth became greater than 6% diameter of FRP bars, corresponding to a strength retention of 77.4%. Consequently, the bond strength between FRP bars and concrete can no longer be guaranteed.

• The proposed FRP bar-to-concrete bond failure criteria define a quantitative parameter to evaluate the bonding conditions of FRP bar-concrete structures as service time increases, providing insights and new perspective to the current design guides. It should be noted that the 6% criterion is derived from limited rib-geometry reported in Zhang et al. (2024b,c). Future work is needed to derive a more general criterion to cover different surface treatments, embedment lengths, or concrete strengths.

Footnotes

Acknowledgements

We sincerely appreciate all the valuable comments and suggestions from the reviewers, which helped us in improving the quality of the manuscript. The manuscript has been carefully revised according to these suggestions and comments. A detailed response to the comments point-by-point is attached. For the sake of clarity, the Reviewers’ comments are shown in ‘black’, while the authors’ responses in ‘blue’. The changes are shown in red text in the revised manuscript.

ORCID iD

Xiao-Ling Zhao

Author contribution

Qi Zhao: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Validation, Visualization, Writing – original draft.

Daxu Zhang: Conceptualization, Methodology, Writing – review & editing.

Jie Liu: Methodology, Writing – review & editing.

Keitai Iwama: Methodology, Writing – review & editing.

Pei-Fu Zhang: Methodology, Writing – review & editing.

Lingxin Zeng: Methodology, Writing – review & editing.

Xiao-Ling Zhao: Conceptualization, Methodology, Project administration, Resources, Supervision, Writing – review & editing.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The first author is sponsored by the Innovation and Technology Fund (ITF) Research Talent Hub of Hong Kong. The study is supported by the Henan Province Science and Technology Research Projects (242102320013).

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.

Declaration of generative AI in scientific writing

The authors declare that any generative AI was not used in preparation of the work submitted.

Data Availability Statement

Data will be made available on request from corresponding author.*

Appendix

Table A1.

Diffusion Coefficients of FRP Composites Used in the Validation.

Data source	FRP type	Exposure environment	Temperature	Diffusion coefficient(mm²/s)
(Wu et al., 2015)	BFRP bars	Concrete pore solution with a pH Value of 13.0	25°C	5.26×10^-11
			40°C	2.49×10^-10
			55°C	1.02×10^-9
(Wang et al., 2017b)	BFRP bars	Seawater sea-sand concrete pore solution with a pH value of 13.4	32°C	9.86×10^-11
			40°C	4.46×10^-10
			48°C	1.87×10^-9
			55°C	6.20×10^-9
(Wang et al., 2017b)	BFRP bars	Seawater sea-sand concrete pore solution with a pH value of 12.7	32°C	1.06×10^-10
			40°C	2.06×10^-10
			55°C	6.52×10^-10
(Wang et al., 2017a)	CFRP bars	Seawater sea-sand concrete pore solution with a pH value of 13.4	25°C	6.26×10^-11
			40°C	5.36×10^-10
			55°C	3.77×10^-9
(Wang et al., 2017b)	CFRP bars	Seawater sea-sand concrete pore solution with a pH value of 12.7	25°C	5.56×10^-11
			40°C	2.89×10^-10
			55°C	1.29×10^-9
(Wang et al., 2017a)	GFRP bars	Seawater sea-sand concrete pore solution with a pH value of 13.4	25°C	8.65×10^-11
			40°C	4.21×10^-10
			55°C	1.77×10^-9
(Wang et al., 2017b)	GFRP bars	Seawater sea-sand concrete pore solution with a pH value of 12.7	25°C	5.84×10^-11
			40°C	1.36×10^-10
			55°C	2.93×10^-10
(Wang et al., 2020)	BFRP laminates	Alkaline solution with a pH value Of 13.0	60°C	3.60×10^-9
(Bazli et al., 2020b)	GFRP tubes	Seawater sea-sand concrete pore solution with a pH value of 13.4	25°C	2.5×10^-9
(Bazli et al., 2020b)	GFRP tubes		60°C	9.6×10^-9

References

ACI Committee 440 (2015) Guide for the Design and Construction of Concrete Reinforced with FRP Bars (ACI 440.1R-15). Farmington Hills, MI: American Concrete Institute.

ACI Committee 440 (2017) Guide for the Design and Construction of Externally Bonded FRP Systems for Strengthening Concrete Structures (ACI 440.2R-17). Farmington Hill, MI: American Concrete Institute.

ACI Committee 440 (2023) Building Code Requirements for Structural Concrete Reinforced with Glass Fiber-Reinforced Polymer (GFRP) Bars-Code and Commentary ACI 440. Farmington Hills, MI: American Concrete Institute, 11–22.

Allen

Atadero

(2012) Evaluating the long-term durability of externally bonded FRP via field assessments. Journal of Composites for Construction 16(6): 737–746.

Antoon

Koenig

(1980) The structure and moisture stability of the matrix phase in glass-reinforced epoxy composites. Journal of Macromolecular Science, Part C 19(1): 135–173.

Arias

JPM

Bernal

Vázquez

, et al. (2018) Aging in water and in an alkaline medium of unsaturated polyester and epoxy resins: experimental study and modeling. Advances in Polymer Technology 37(2): 450–460.

Bazli

Zhao

Bai

, et al. (2020a) Durability of pultruded GFRP tubes subjected to seawater sea sand concrete and seawater environments. Construction and Building Materials 245: 118399.

Bazli

Zhao

Jafari

, et al. (2020b) Mechanical properties of pultruded GFRP profiles under seawater sea sand concrete environment coupled with UV radiation and moisture. Construction and Building Materials 258: 120369.

Bazli

Zhao

Jafari

, et al. (2021) Durability of glass-fibre-reinforced polymer composites under seawater and sea-sand concrete coupled with harsh outdoor environments. Advances in Structural Engineering 24(6): 1090–1109.

10.

Benmokrane

Hassan

Robert

, et al. (2020a) Effect of different constituent fiber, resin, and sizing combinations on alkaline resistance of basalt, carbon, and glass FRP bars. Journal of Composites for Construction 24(3): 1–18.

11.

Benmokrane

Brown

Ali

, et al. (2020b) Reconsideration of the environmental reduction factor CE for GFRP reinforcing bars in concrete structures. Journal of Composites for Construction 24(4): 1–6.

12.

Brand

Veith

Baier

, et al. (2020) New methodical approaches for the investigation of weathered epoxy resins used for corrosion protection of steel constructions. Journal of Hazardous Materials 395: 122289.

13.

CEN/TS 19101 (2022) Design of fibre-polymer composite structures.

14.

Ceroni

Bonati

Galimberti

, et al. (2018) Effects of environmental conditioning on the bond behavior of FRP and FRCM systems applied to concrete elements. Journal of Engineering Mechanics 144(1): 1–15.

15.

Chen

Davalos

Ray

(2006) Durability prediction for GFRP reinforcing bars using short-term data of accelerated aging tests. Journal of Composites for Construction 10(4): 279–286.

16.

Chen

Davalos

Ray

, et al. (2007) Accelerated aging tests for evaluations of durability performance of FRP reinforcing bars for concrete structures. Composite Structures 78(1): 101–111.

17.

Chin

Aouadi

Haight

, et al. (2001) Effects of water, salt solution and simulated concrete pore solution on the properties of composite matrix resins used in civil engineering applications. Polymer Composites 22(2): 282–297.

18.

Correia

Keller

Garrido

, et al. (2023) Mechanical properties of FRP materials at elevated temperature – definition of a temperature conversion factor for design in service conditions. Construction and Building Materials 367: 130298.

19.

Dhondy

Xiang

, et al. (2021) Effects of mixing water salinity on the properties of concrete. Advances in Structural Engineering 24(6): 1150–1160.

20.

Zhang

, et al. (2024) Predicting tensile behaviour of plain weave CMCs using a nonlinear data-driven constitutive model for fibre tow composites. Ceramics International 50(9): 16142–16154.

21.

Fang

Guo

(2023) Deterioration mechanism of hydrothermal aging on the properties of carbon fiber/epoxy composites in various media. Journal of Composite Materials 57(7): 1235–1246.

22.

Gao

Wang

Zhou

, et al. (2020) The performance of epoxy coatings containing polyaniline (PANI) nanowires in neutral salt, alkaline, and acidic aqueous media. Journal of Applied Polymer Science 137(36): 49049.

23.

Gooranorimi

Nanni

(2017) GFRP reinforcement in concrete after 15 years of service. Journal of Composites for Construction 21(5): 04017024.

24.

Guo

Al-Saadi

Singh Raman

, et al. (2022) Durability of fibre reinforced polymers in exposure to dual environment of seawater sea sand concrete and seawater. Materials 15(14): 4967.

25.

Hojo

Tsuda

Ogasawara

(1991) Form and rate of corrosion of corrosion-resistant FRP resins. Advanced Composite Materials 1(1): 55–67.

26.

Hojo

Tsuda

Kubouchi

, et al. (1998) Corrosion of plastics and composites in chemical environments. Metals and Materials International 4(6): 1191–1197.

27.

Huang

Aboutaha

(2010) Environmental reduction factors for GFRP bars used as concrete reinforcement: new scientific approach. Journal of Composites for Construction 14(5): 479–486.

28.

Kaushik

Islam

(1995) Suitability of sea water for mixing structural concrete exposed to a marine environment. Cement and Concrete Composites 17(3): 177–185.

29.

Kootsookos

Mouritz

(2004) Seawater durability of glass- and carbon-polymer composites. Composites Science and Technology 64(10–11): 1503–1511.

30.

Xie

Liu

, et al. (2019) A critical review and assessment for FRP-Concrete bond systems with epoxy resin exposed to chloride environments. Composite Structures 229: 111372.

31.

Zhao

Raman

(2021) Durability of seawater and sea sand concrete and seawater and sea sand concrete–filled fibre-reinforced polymer/stainless steel tubular stub columns. Advances in Structural Engineering 24(6): 1074–1089.

32.

Liao

Zeng

Bai

, et al. (2022) Bond strength of GFRP bars to high strength and ultra-high strength fiber reinforced seawater sea-sand concrete (SSC). Composite Structures 281: 115013.

33.

Lim

JSK

Gan

(2019) Unraveling the mechanistic origins of epoxy degradation in acids. ACS Omega 4(6): 10799–10808.

34.

LiuHu

Wang

Zhao

, et al. (2022) Numerical study on cracking and its effect on chloride transport in concrete subjected to external load. Construction and Building Materials 325: 126797.

35.

Liu

Wei

Zhao

, et al. (2024) Fatigue-induced fracture assessment for FRP-Steel bonding joints after seawater exposure. Structures 68: 107209.

36.

Long

Bai

Liu

, et al. (2023) Association between serviceability and recyclability of amine-cured epoxy thermosets. ACS Sustainable Chemistry & Engineering 11(34): 12790–12797.

37.

Xie

(2021) Durability of BFRP bars wrapped in seawater sea sand concrete. Composite Structures 255(100): 112935.

38.

Kim

Nutt

(2017) Chemical treatment for dissolution of amine-cured epoxies at atmospheric pressure. Polymer Degradation and Stability 146: 240–249.

39.

Mallick

(2018) Processing of Polymer Matrix Composites. Boca Raton: CRC Press.

40.

Mehta

Monteiro

PJM

(2014) Concrete: Microstructure, Properties, and Materials. 4th edition. New York: McGraw-Hill Education.

41.

Myers

Viswanath

(2006) A worldwide survey of environmental reduction factors for fiber reinforced polymers (FRP). In: Proceedings of the Structures Congress and Exposition 2006, 96.

42.

Ortiz

Khedmatgozar Dolati

Malla

, et al. (2023) FRP-Reinforced/Strengthened concrete: state-Of-The-art review on durability and mechanical effects. Materials 16(5): 1990.

43.

Peebles

(2018) Carbon Fibers: Formation, Structure, and Properties. Taylor and Francis Press.

44.

Sembokuya

Negishi

Kubouchi

, et al. (2003) Corrosion behavior of epoxy resin cured with different amount of hardener in corrosive solutions. Materials Science Research International 52(3): 230–234.

45.

Shrestha

Ueda

Zhang

(2015) Durability of FRP concrete bonds and its constituent properties under the influence of moisture conditions. Journal of Materials in Civil Engineering 27(2): 1–14.

46.

Sun

Zheng

Zhou

, et al. (2021) A study of the bond behavior of FRP bars in MPC seawater concrete. Advances in Structural Engineering 24(6): 1110–1123.

47.

Taha

Alnahhal

(2021) Bond durability and service life prediction of BFRP bars to steel FRC under aggressive environmental conditions. Composite Structures 269: 114034.

48.

Tanks

Arao

Kubouchi

(2022) Network-level analysis of damage in amine-crosslinked diglycidyl ether resins degraded by acid. Express Polymer Letters 16(5): 488–499.

49.

Teng

Xiang

, et al. (2019) Development and mechanical behaviour of ultra-high-performance seawater sea-sand concrete. Advances in Structural Engineering 22(14): 3100–3120.

50.

Wang

Zhao

Xian

, et al. (2017a) Durability study on interlaminar shear behaviour of basalt-glass- and carbon-fibre reinforced polymer (B/G/CFRP) bars in seawater sea sand concrete environment. Construction and Building Materials 156: 985–1004.

51.

Wang

Zhao

Xian

, et al. (2017b) Long-term durability of basalt- and glass-fibre reinforced polymer (BFRP/GFRP) bars in seawater and sea sand concrete environment. Construction and Building Materials 139: 467–489.

52.

Wang

Zhu

Zhang

, et al. (2020) Influence of thickness on water absorption and tensile strength of BFRP laminates in water or alkaline solution and a thickness-dependent accelerated ageing method for BFRP laminates. Applied Sciences (Switzerland) 10(10): 3618.

53.

Dong

Z-Q

Wang

, et al. (2015) Prediction of long-term performance and durability of BFRP bars under the combined effect of sustained load and corrosive solutions. Journal of Composites for Construction 19(3): 04014058.

54.

Yang

Xie

, et al. (2022) Mechanical properties and environmental performance of seawater sea-sand self-compacting concrete. Advances in Structural Engineering 25(15): 3114–3136.

55.

Zhang

Xiao

(2022a) Durability of FRP bars and FRP bar reinforced seawater sea sand concrete structures in marine environments. Construction and Building Materials 350: 128898.

56.

Zhang

Xiao

Hou

, et al. (2022b) Experimental study on carbonation behavior of seawater sea sand recycled aggregate concrete. Advances in Structural Engineering 25(5): 927–938.

57.

Zhang

P-F

Zhang

ZhaoZhao

, et al. (2024a) Natural language processing‐based deep transfer learning model across diverse tabular datasets for bond strength prediction of composite bars in concrete. Computer-Aided Civil and Infrastructure Engineering 40: 917–939.

58.

Zhang

Iqbal

Zhang

, et al. (2024b) Bond strength prediction of FRP bars to seawater sea sand concrete based on ensemble learning models. Engineering Structures 302: 117382.

59.

Zhang

Zhao

Zhang

, et al. (2024c) Prediction of bond strength and failure mode of FRP bars embedded in UHPC or UHPSSC utilising extreme gradient boosting technique. Composite Structures 346: 118437.

60.

Zhao

Dang

J-TT

, et al. (2019) Experimental investigation of shear walls using carbon fiber reinforced polymer bars under cyclic lateral loading. Engineering Structures 191: 82–91.

61.

Zhao

Zhang

Zhao

X-L

, et al. (2021) Modelling damage evolution of carbon fiber-reinforced epoxy polymer composites in seawater sea sand concrete environment. Composites Science and Technology 215: 108961.

62.

Zhao

Zhang

Zhao

, et al. (2022) Experimental and modelling studies on damage evolution of epoxy-based GFRP bars in pore solution environment of seawater sea-sand concrete. China Civil Engineering Journal 55(9): 25–41.

63.

Zhao

Zhang

, et al. (2024a) Degradation of GFRP bars with epoxy and vinyl ester matrices in a marine concrete environment: an experimental study and theoretical modeling. Journal of Composites for Construction 28(2): 04024004.

64.

Zhao

Iwama

Dai

, et al. (2024b) Deterioration modelling of GFRP-Reinforced cement-based concrete infrastructure in service under the natural inland atmospheric environment. Construction and Building Materials 447: 138005.

65.

Zhao

X-L

Zhang

, et al. (2024c) Effects of exposure in seawater sea-sand concrete pore solution on fatigue performance of carbon FRP bars. Composites Science and Technology 247: 110418.

66.

Zhao

Zhang

Zhao

, et al. (2025a) A SHAP algorithm-based prediction of the interlaminar shear strength degradation of G/BFRP bars embedded in concrete exposed to marine environment. Case Studies in Construction Materials 22: e04770.

67.

Zhao

Zhang

, et al. (2025b) Prediction of interlaminar shear strength retention of FRP bars in marine concrete environments using XGBoost model. Journal of Building Engineering 105: 112466.

68.

Zhou

Feng

Yang

(2021) Advances in coral aggregate concrete and its combination with FRP: a state-of-the-art review. Advances in Structural Engineering 24(6): 1161–1181.