To determine the microstructure and mechanical properties of corrosion products in marine environment, samples were collected and analyzed based on the Observation and Research Base of Transport Industry of Long-term Performances of Marine Infrastructure (the Field Observation Base) and cracked beams in Zhejiang province. Specifically, sample A and sample B were derived from steel plates and concrete specimens with 10 mm and 20 mm covers in the Field Observation Base, respectively. Sample C was from the cracked beams in Zhejiang province. In addition, the steel plates, concrete specimens and cracked beams were exposed in the splash zone and the steel type employed was HRB300 with ASTM A615 Grade 60. The basic information of corrosion products is shown in Table 1. The sources and samples of corrosion products are depicted in Figures 1, 2, respectively.

The XRD patterns of the corrosion products of steel are shown in Figure 5. Only the peaks related to the steel corrosion products are labelled. It can be observed that the main components of corrosion products from steel plates include α-FeOOH, γ-FeOOH, Fe₃O₄, and Fe₂O₃. The main components of corrosion products from the concrete specimens and cracked beams include α-FeOOH, Fe₃O₄, and Fe₂O₃. γ-FeOOH is a transitional hydroxyl oxide of the more stable α-FeOOH (Wang et al., 2019; Gu et al., 2019; Duffó et al., 2004; Feng et al., 2015). Additionally, SiO₂ peaks appear in the corrosion products from the concrete specimens and cracked beams, but generally disappear in the steel plate corrosion products. This might be attributed to the penetration of soluble cement paste from concrete into the corrosion products. In conclusion, the main components of corrosion products are oxides and hydroxyl oxides of iron.

The mass ratios of corrosion product components were determined by TG. And the obtained TG and Derivative thermal gravimetric (DTG) curves of corrosion product are depicted in Figure 6. In the DTG curve, the peak at 30°C–200°C represents the water evaporation while the peak at 200°C–500°C corresponds to the dehydration of FeOOH, as shown in Equation 4 (Xu et al., 2017; Melchers, 2023). 2 F e O O H → F e 2 O 3 + H 2 O (4)

With reference to the corresponding TG curves, the mass loss due to the dehydration of FeOOH can be obtained, and the quantity of oxides and hydroxyl hydroxides in the corrosion products can be determined, as shown in Figure 7. It can be seen that the mass ratio of hydroxyl oxides in B₁, B₂, C, and A increases in order, while the mass ratio of oxides decreases in order correspondingly.

Based on the XRD analysis, it can be seen that the main components of corrosion products include oxides and hydroxyl oxides of iron. To determine the volume expansion coefficient of the corrosion products, it is necessary to calculate the volume expansion coefficients of each corrosion product component. These coefficients can be measured based on the density and molar mass of the components.

For example, the density and molar mass of Fe₂O₃ are 5.25 g/cm³ and 160 g/mol, respectively.

Since 1 mol of Fe can be oxidized to 1/2 mol of Fe₂O₃, the mass of the corrosion product increases to 1.43 times the original mass of Fe. The volume expansion coefficient of Fe₂O₃ can be determined by Equation 5. Similarly, the volume expansion coefficients of other corrosion product components can be obtained, as shown in Figure 8. η F e 2 O 3 = V F e 2 O 3 V F e = 1.43 m s / ρ F e 2 O 3 m s / ρ F e = 2.15 (5)

From Figure 8, it can be seen that the volume expansion coefficients of Fe₃O₄ and Fe₂O₃, α-FeOOH and γ-FeOOH are close to each other. The average volume expansion coefficients of Fe₃O₄ and Fe₂O₃, α-FeOOH and γ-FeOOH can be used to represent the volume expansion coefficients of oxides and hydroxyl oxides in corrosion products, respectively. Therefore, the volume expansion coefficients of corrosion products can be determined according to Equation 6, as shown in Figure 9. n = a × η 1 + b × η 2 (6)

Where n represents the volume expansion coefficient of corrosion product; a and b represent the mass ratios of oxide and hydroxyl oxide in the corrosion products, respectively; η ₁ and η ₂ represent the volume expansion coefficients of oxides and hydroxyl oxides in the corrosion products, respectively.

From Figure 9, it can be seen that the volume expansion coefficients of B₁, B₂, C and A increase in order to 2.19,2.51, 2.59 and 2.68, respectively.

The SEM images and EDS analysis results of the corrosion products are depicted in Figures 10, 11, respectively. It can be observed from Figure 10 that the corrosion products are generally rough. Corrosion products from the cracked beams appear looser and exhibit some cracking compared to those from concrete specimens. In addition, corrosion products from the steel plates are noted to be loose and porous.

Figure 11 shows that the Fe/O ratio of corrosion products from sample B₁, B₂, C, and A decreases in order. This indicates that the oxygen content of the corrosion products increases in the following order: concrete specimens with 20 mm cover, concrete specimens with 10 mm cover, cracked beams, and steel plates. This trend suggests that the oxidation degrees of these corrosion products increase correspondingly.

The EDS analysis results align with the TG analysis, reinforcing the observations made.

The nanoindentation experiment was conducted to determine the mechanical properties of the corrosion products. Six indentation points were used for each corrosion product sample, and the obtained stress-strain curves are shown in Figure 12. It can be observed from Figure 12 that the stress-strain curve shapes of different indentation points for corrosion products from steel plates are consistent, indicating that these corrosion products are uniform. Additionally, with the exception of indentation point 5 in sample B₁ and indentation point 4 in sample B₂, the stress-strain curve shapes for corrosion products from concrete specimens and cracked beams are similar. This similarity demonstrates that the corrosion products from concrete in marine environments are stable and uniform. The obtained nanoindentation stress-strain curves confirm that nanoindentation experiments are a feasible method for measuring the mechanical properties of corrosion products.

The nanoindentation elastic modulus and hardness of different indentation points for corrosion products are shown in Table 2; Figure 13; Table 3; Figure 14, respectively. It can be observed from Table 2; Figure 13 that the average elastic modulus for samples A, B₁, B₂, and C are 82.0 GPa, 110.9 GPa, 100.4 GPa, and 83.0 GPa, respectively. Additionally, Table 3; Figure 14 indicate that the average hardness of samples A, B₁, B₂, and C are 0.011 GPa, 0.008 GPa, 0.006 GPa, and 0.006 GPa, respectively.

The variation in elastic modulus of corrosion products is attributed to different corrosion rates and corrosion spaces. Specifically, concrete specimens are small and less affected by loading and cracking. The steel corrosion rate in these specimens is low. Furthermore, the limited corrosion space results in high corrosion expansion force, leading to stable, dense corrosion products with a higher elastic modulus.

In contrast, the cracked beams have several cracks, providing ample oxygen and space for the steel corrosion process. Consequently, the corrosion rate is relatively high, and the corrosion expansion force is relatively low, resulting in looser corrosion products with a lower elastic modulus compared to those from concrete specimens. For steel plates without concrete cover, the corrosion rate is high, and the corrosion space is sufficient, indicating a lower corrosion expansion force. Thus, the corrosion products from steel plates are looser and exhibit a lower elastic modulus compared to those from other sources.

In conclusion, the corrosion rate of concrete structures in marine environments is low, and the corrosion space is limited. Additionally, the high corrosion expansion force leads to dense corrosion products. The elastic modulus of corrosion products from concrete ranges from 70 to 110 GPa, while the hardness ranges from 0.002 to 0.013 GPa.

The pressure and height (P-h) curves of corrosion products obtained from the standard consolidation test are shown in Figure 15. During the measurement, graded pressures were applied, and the load remained stable at each grade. When the pressure was stable, the corrosion products could be approximated as linear elastic bodies and the instantaneous elastic modulus can be obtained. The instantaneous elastic modulus of the corrosion products was calculated using Hertz elastic contact theory, as shown in Equations 7, 8 (Sorelli et al., 2008). δ = 9 F g 2 16 r E * 2 1 3 (7) E = 9 F g 2 2 − 2 v 2 2 8 r δ 3 (8)

Where Fg is the vertical force applied to the corrosion product particle, r is the equivalent radius, E* is the equivalent elastic modulus, δ is the vertical deformation, E is the instantaneous elastic modulus, and v (0.25) is the Poisson’s ratio (Xu et al., 2014).

The instantaneous elastic modulus of samples A and B₁ is shown in Figure 16. It can be observed that the corrosion products do not exhibit purely linear elastic behavior. When the load exceeds 100 kPa, the instantaneous elastic modulus of samples A and B₁ increases with the applied pressure. Additionally, the instantaneous elastic modulus of sample B₁ is higher than that of sample A, with this trend being most pronounced at pressures ranging from 50 to 400 kPa. The difference in instantaneous elastic modulus between samples A and B₁ decreases with increasing pressure. At a pressure of 1,600 kPa, the instantaneous elastic modulus of samples A and B₁ are 149.0 GPa and 127.9 GPa, respectively.