Material that is used in this study is GFRP that consists of glass fiber and matrix resin. Glass fiber reinforcement used in this study is chopped-strand mat E-glass type. Vinyl ester resin made from epoxy novolac is used for chemical storage tanks, pipes, ducting, fume extraction systems, and gas cleaning units because it can resist chemicals well even at high temperatures. These exhibit tensile elongation and superior corrosion resistance, rendering them a potential material for manufacturing lining coatings with exceptional adherence to various polymers and traditional materials such as steel and concrete 23. Vinyl ester resin DERAKENE MFE 711 (produced by ASHLAND) was used in this study. Methyl Ethyl Ketone Peroxide (MEKP) produced by PT. Kawaguchi Kimia Indonesia was used as a catalyst and curing agent. The 6% cobalt naphthenate (produced by PT. Justus Sakti Raya) was used as a accelerant. Chopped-strand mat E-glass type fiber was used as the reinforcing fiber.

Epoxy resin is also used in this study as a comparison. Epoxy is increasingly utilized due to its high specific stiffness and strength, corrosion resistance, and chemical compatibility with reinforcing fibers 2425. Bisphenol A-diglycidyl ether (DGEBA, R140) epoxy resin was used in this study. Poly(oxypropylene) diamine JEFFAMINE D230 was used in this study as curing agent.

Vinyl ester GFRP specimens were prepared by mixing the resin with curing agent and catalyst at a mass ratio of 100:1:0.5, respectively. After mixing and degassing process, the mixture was poured into the glass fiber on the top steel mold and cured at room temperature for 24 hours. The cured glass fiber-VE plates were then removed from the mold. The specimen was precisely cut into small pieces at a dimension of 60 × 25 × 2 mm (length × width × thickness), as shown in Figure 1. To remove mold release and other material that could interfere with surface adsorption, all specimens were cleaned with water, then dried at room temperature for 1-2 days.

Specimen of epoxy was prepared by mixing with the curing agent at a mass ratio of 100:30.6. After mixing and degassing process, it was poured into the glass fiber on pre-heated steel mold and cured using a hot press at a schedule of 60˚C for 6 hours for the first curing and 110 ˚C for 12 hours for the second curing. After finishing the curing procedure in a hot press, cooling to room temperature, and the cured epoxy plates were removed from the mold.

Vinyl ester GFRP and epoxy GFRP specimens were prepared for both control and cyclic concentration exposure tests. The control specimens were fabricated using the same mixed design and manufacturing process as the test specimens but were not subjected to any cyclic exposure. Instead, the baseline material behavior under neutral conditions for comparison.

The specimens were prepared for both vapor phase and liquid phase. Specimens were exposed to varying HCl concentrations at specified time intervals. Specimens are changed from higher HCl concentration to lower concentration or water alternately. In this study, for vinyl ester specimens, six pairs of HCl concentrations were tested to see how they break down when switched between different levels: 35 mass% HCl and water, 20 mass% HCl and water, 10 mass% HCl and water. Epoxy specimens are only exposed to 35 mass% HCl and 0 mass% HCl (water). The temperature was constantly kept at 40˚C. For HCl exposure, experiment was done inside an acid chamber. Meanwhile, for water exposure, the specimens in glass beakers filled with water were placed in a 40˚C water bath.

The specimen undergoes a full exposure cycle of 720 hours (30 days). A 5-day exposure in a solution is classified as half-cycles; thus, specimens initially exposed to HCl and subsequently to water represent one complete cycle, as illustrated in Figure 2. Schemes 2 and 3 were implemented to examine the variation in degradation resulting from exposure to varying concentrations of HCl cycles. The experiment included HCl concentrations of 35 mass%, 20 mass%, and 10 mass%, as illustrated in Figure 3 and Figure 4.

Gravimetric analysis was used to estimate the amount of penetrated acid solutions by measuring mass change and dimensional change. Each specimen was removed from the solution and patted dried by towel/tissue, and then the mass was measured on a digital balance with an accuracy of 0.0001 g. The change in mass of the specimens was calculated based on Eq. 1, where M_t is mass of specimen at time t and M₀ is initial mass of specimen.

(1)

Dimensional measurements were done using digital caliper with an accuracy of 1μm. The changes in thickness, length, and width were calculated based on Eq. (2-4).

Where h is thickness of specimen at time t, h₀ is initial thickness of specimen, l_t is length of specimen at time t, l₀ is initial length of specimen, w_tis width of specimen at time t, and w₀ is initial width of specimen.

A three-point bending test was conducted using the Shimadzu Autograph AGS-1KNJ machine to assess the impact of acid penetration on the mechanical properties of the specimens, specifically to determine the flexural strength of the material following exposure to the solution. The test was performed at a crosshead speed of 2 mm/min, following ASTM D790 standards 26.

Following the bending test, the cross-section of the specimen exposed to the vapor phase and liquid phase was observed using JEOL JSM-6510LA SEM/EDS machine. SEM was used to examine the morphology of fracture surfaces. Meanwhile, EDS was used to determine the chemical composition of materials and to create element composition maps, which was used to determine the penetration depth of Cl in the specimen.

FTIR observation is conducted to determine chemical reactions that occurred on cured VE while exposed to cyclic concentrations. FTIR spectra were obtained using an FTIR-ATR (Shimadzu AIM-8000R) at a resolution of 4 cm^-1 and a minimum of 60 scans averaged per spectrum. To get a representation of the pattern of the whole surface area, spectra from at least three points on a specimen were averaged.

SEC is a technique that separates molecules based on their sizes, allowing for the determination of the molecular weight distribution of the organic soluble components of a polymer 2728. Through SEC analysis, it can be determined whether the resin is leaching out in the HCl solution. The sample preparation for SEC analysis is as shown in Figure 5. After exposing GFRP to HCl solution for 30 days, the post-exposure HCl solution was collected for analysis. The HCl solution was subjected to liquid-liquid extraction using ethyl acetate. Following extraction, the mixture is separated into an organic phase (resin dissolved in ethyl acetate) and an aqueous phase (HCl and ethyl acetate). To neutralize any residual acidity, the organic phase was treated with a saturated aqueous solution of sodium bicarbonate (NaHCO₃), and the resulting aqueous layer was removed. The purified organic phase was then concentrated by evaporation under reduced pressure to remove ethyl acetate, yielding an organic solid. This solid was subsequently dissolved in tetrahydrofuran (THF) at a concentration of 100 mg per 10 ml. Finally, a 0.4 ml (400 µl) aliquot of the THF solution was injected into a SEC column for molecular weight distribution analysis.

The mass uptake was analyzed using gravimetric analysis following exposure of the specimens to cyclic concentration. The result of mass uptake measurement after being exposed to cyclic concentration of 35 mass% HCl and water for 720 hours are shown in Figure 6. In Cycle 0.5 (0-120 h) penetration occurred on the surface of the GFRP composite when first exposed to 35 mass% HCl. Between 120 and 240 hours, the specimens are exposed to water, during which water absorption primarily occurs due to osmotic pressure. This pressure drives water from the surrounding environment into the polymer matrix, leading to an increase in mass uptake observed during cycle 1 (120–240 h). Under liquid-phase conditions, HCl is initially concentrated at the surface of the polymer after initially exposed to a 35 mass% HCl. Additionally, a relatively large volume of water is present, leading to a high concentration difference of HCl between the polymer surface and surrounding solution. This promotes the leaching of HCl from the polymer into the surrounding solution. In contrast, under vapor-phase conditions, the amount of water available is significantly lower. Water condenses on the polymer surface, forming only a thin layer. This limited water availability slows the diffusion of HCl out from the polymer. Moreover, the thin condensed water layer may become saturated with HCl, further hindering its release into the vapor phase. As a result of these mechanisms, the polymer in the vapor-phase condition exhibits a greater mass uptake than in the liquid-phase condition, despite the lower water content.

During Cycle 1.5 (240–360 h), both the liquid and vapor phase specimens are re-exposed to the HCl solution, allowing them to encounter the acid once again. By this point, a significant amount of water has already been absorbed into the polymer. When the specimens are placed back into a high-concentration HCl environment, the external solution draws water out of the polymer due to osmotic effects 29303132. As a result, the mass of specimens decreases during this cycle. When exposed to vapor, a significant amount of water is trapped inside the polymer. This trapped water may help the polymer return to its previous level of mass uptake. However, since the duration of Cycle 1.5 (240–360 h) is relatively short, the mass uptake in the vapor phase remains slightly higher than in the liquid phase during this period. This suggests that the same mechanisms observed earlier such as limited leaching and high internal water content may continue to drive mass uptake in the vapor phase. Cycle 2 and 3 follow the same mechanism observed in Cycle 1, where the specimens absorb water during exposure, primarily driven by osmotic pressure. Similarly, Cycle 2.5 (480–600 h) exhibits the same behavior as Cycle 1.5 (240–360 h), during which the specimens are re-exposed to the HCl solution, resulting in water being drawn out of the polymer and leading to a decrease in mass uptake.

Figure 7 (a) and Figure 7 (b) present a comparison of mass uptake for specimens exposed to cyclic concentrations of 35 mass% HCl-water, 20 mass% HCl-water, and 10 mass% HCl-water. The specimen exposed to 35 mass% HCl-water shows a significantly higher mass uptake compared to those exposed to 10 mass% HCl-water and 20 mass% HCl-water. One possible reason for this difference is that 10 mass% HCl-water and 20 mass% HCl-water are far from the maximum achievable HCl concentration under normal atmospheric pressure. This lower concentration may result in weaker osmotic driving forces and less aggressive diffusion behavior, which could influence the overall absorption mechanism.

In addition to determining the mass uptake, dimensional change was also assessed. Figure 8 illustrates the changes in thickness, length, and width for both the vapor phase and liquid phase. The results indicate that the increase in thickness is significantly larger than the changes in length and width. This finding suggests that there is minimal change in the size of the specimen, but there may be surface penetration. In GFRP specimens, dimensional changes during environmental exposure are often anisotropic (vary depending on the direction) due to the material’s internal structure 3334. GFRP is a composite made of stiff, aligned glass fibers embedded in a polymer matrix. The glass fibers in the specimen used for this research are oriented in the plane of the specimen, primarily along the length and width directions, which provides high stiffness and dimensional stability in those directions 3536 As a result, any swelling or dimensional change due to mass uptake is largely constrained along the length and width because the fibers restrict expansion. In contrast, the thickness direction contains fewer or no reinforcing fibers, making it less constrained. Therefore, when water is absorbed into the polymer matrix, the matrix can swell more freely in the thickness direction, leading to measurable dimensional change in thickness, while the length and width remain relatively unchanged.

Figure 8 shows that the thickness change in vapor phase is bigger than liquid phase. This result suggests that similar mechanisms are at play in both conditions, and the changes are proportional to the mass uptake observed. The phenomenon of osmotic pressure induces the swelling of materials and subsequently contributes to their degradation.

The mechanical properties of the vapor phase and liquid phase specimens were analyzed using a three-point bending test. The flexural strength of the specimens exposed to cyclic concentration conditions can be seen in Figure 9. Acid penetration results in decreased mechanical properties. As shown in Sections 3.1 and 3.2, HCl and water penetrate the network and cause significant swelling, leading to plasticization. Moreover, this plasticization/swelling effect was the primary reason for the decrease in flexural strength 3738. If swelling occurs, very short molecular weight chemicals inserted in the matrix which resulted in reduced polymer-polymer interaction and increased plasticization 39 Swelling does not result in direct interaction, but 3D construction will be extended, which resulted in plasticization.

The flexural strength of the vapor phase and liquid phase specimens show different trends. When comparing the flexural strength in the vapor phase, there is a similarity in the results between exposure to 10 mass% HCl-water and 20 mass% HCl-water. However, there is a significant difference in flexural strength after exposure to 35 mass% HCl-water. Meanwhile, in the liquid phase, the flexural strength values for all exposure conditions showed similar results.

Following the bending test, the specimen cross-section surface in the vapor phase and liquid phase was observed and analyzed. The penetration depth of Cl from mapping of cross-section surface is investigated for each vapor phase and liquid phase exposure at initial condition and after 5 days, 10 days, 15 days, and 20 days of 35 mass% HCl-water cyclic exposure.

The progression of Cl penetration with immersion time was confirmed. Figure 10 shows that Cl penetration in VE occurs at a very slow rate compared to epoxy (EP) 40414243. There is no significant difference in the depth of Cl penetration between exposure to the vapor and liquid phases. This small difference is attributed to the mechanisms of Cl infiltration. In the vapor phase, Cl enters the material either as a gas or as condensed Cl, leading to a lower overall concentration compared to the direct dissolution and transport in the liquid phase. Additionally, the Cl distribution appears concentrated near the surface, deviating from the expected behavior based on simple Fick’s law diffusion 4445. This deviation is likely due to the effects of cyclic concentration processes near the surface. While the distribution pattern of Cl does not vary significantly between phases, the key distinction lies in water uptake behavior. The amount of water absorbed differs greatly between the vapor and liquid environments, which is the primary reason for the observed discrepancy in Cl accumulation.

As depicted in Figure 11, the progression of hydrolysis was confirmed at a preliminary level through infrared spectroscopy, wherein a decrease in the ester band (more specifically a combination of C–O stretch and O–H bend) is observed 46. The peaks at 1605 cm⁻¹, 1509 cm⁻¹, and 1452 cm⁻¹ (benzene ring) and at 1240 cm⁻¹ and 1041 cm⁻¹ (ether bond) were still present in the products, showing that the carbon framework remained intact during the degradation process. From FTIR results, it can be concluded that hydrolysis has occurred on the material under HCl cyclic concentration exposure.

Components eluted into the hydrochloric acid solution were analyzed to determine whether any resin had leached from the specimen into the solution. The peak observed at a shorter retention time corresponds to high-molecular-weight VE species. In size exclusion chromatography (SEC), larger molecules elute earlier because they are unable to penetrate the porous structure of the column’s packing material. Consequently, these macromolecules are excluded and travel more rapidly through the column, whereas lower molecular weight species enter the pores and elute later 4748495051.

As shown in Figure 12, the first peak at approximately 18 minutes corresponds to low-molecular-weight components such as styrene, methacrylic acid, and catalyst residues. SEC analysis of pure styrene confirmed a single peak at this retention time, supporting this assignment. The second peak, observed at around 35 minutes, is likely associated with the THF solvent itself. Based on these results, estimating the molecular weight of the residual VE resin under cyclic acid exposure is difficult. However, the presence of low-molecular-weight degradation products indicates chain scission at the ends of the vinyl ester backbone, suggesting resin leaching and mass loss from the composite.

The degradation behavior of different resin material in GFRP can be investigated by comparing the evaluation results. In previous research, degradation behavior of GFRP with amine-cured EP was also investigated using gravimetric analysis ⁵². Figure 13 shows how much mass EP and VE resins gain when exposed to HCl and water in both vapor and liquid phase over time. In the case of EP resin, it shows a much bigger mass change than VE resin, particularly in the vapor phase.

In contrast, VE resin shows a relatively stable and gradual increase in mass uptake, especially after transitioning to the water phase. This suggests greater penetration of water or HCl into the resin, but with minimal leaching, as evidenced by the consistent upward trend in mass over time. After switching from HCl to water (around 360 hours), EP resin in both vapor and especially liquid phases shows a clearderease, indicating degradation and material loss. The saturation behavior and negative mass uptake values further support this indicating extensive leaching in the water phase. In the case of VE resin, the mass uptake remains positive throughout the experiment, suggesting less structural degradation. Overall, EP resin is more susceptible to degradation and leaching, while VE resin shows better retention and controlled absorption, indicating higher durability under alternating chemical exposures.

Figure 14 shows the thickness change of different GFRP (EP and VE) over time under HCl-water cyclic concentration in both vapor and liquid phases. In the case of EP, exposure in liquid phase (0-240 hours) leads to a significant increase in thickness, going up by more than 20%. In contrast, the vapor phase (240-480 hours) shows a smaller increase in thickness of 10-15%, meaning there is less swelling or absorption in vapor conditions.

Meanwhile, in the VE case, exposure in vapor phase for 240 to 480 hours leads to a small thickness increase of about 5%, showing it is more resistant than the epoxy resin. In the liquid phase (480-720 hours), the VE shows a minimal thickness change, generally less than 5%, suggesting a stronger resistance to liquid absorption compared to epoxy resin.