Mechanical Performance and Microstructure Evolution in 56-Year-Old Aqueduct Concrete

Abstract

The performance evaluation of aqueducts is crucial for the development of water conservancy and the protection of cultural relics. However, there are few effective methods for accurate evaluations of the mechanical performance of aqueducts. To investigate the changes in the concrete microstructure during the service life of aqueducts, this study conducted compressive tests on various parts of an aqueduct that has been in service for 56 years in Hunan Province, China. Additionally, scanning electron microscopy (SEM) scans and mercury intrusion porosimetry (MIP) tests were carried out on concrete samples taken from the side and bottom of the aqueduct tank. The compressive strength of the aqueduct concrete was 28.3–44.1 MPa, and the porosity of concrete was 10.98–17.57%. The pore structure of concrete is deteriorated by carbonation and water flow, which has a negative impact on the impermeability of the aqueduct. For concrete at the bottom of the tank, the internal pore structure was denser than the external one (with lower porosity and smaller average pore diameter). In contrast, the pore structure in other parts was the opposite. This difference was caused by the presence of flowing water. The types of internal pores in the concrete are basically gel pores and capillary pores. Finally, evaluation models considering the relationships between carbonation, compressive strength, porosity and permeation parameters of aqueduct concrete were proposed. The models can provide theoretical support for the performance evaluation and maintenance of aged aqueducts.

1. Introduction

As an important component of the water conservancy system, aqueducts play a significant role in agricultural irrigation and water resource transportation. China began to build aqueducts in the 1950s [1], and the total number of aqueducts has now surpassed 20,000. The total length of aqueducts in key projects exceeds 500 km, accounting for over 70% of the water conveyance channels in water transfer and core irrigation [2]. Nevertheless, the performance of reinforced concrete structures tends to deteriorate under the coupling influence of loads, environment and water flow scouring [3,4,5,6,7,8,9]. In recent years, many aqueducts have suffered deterioration such as reduced strength, concrete spalling, steel bar corrosion and leakage, which have severely impacted their service performance [10,11,12,13,14]. It is of great significance to research the state of the concrete in aged aqueducts, evaluate and predict the performance for the development of water conservancy and the protection of cultural relics.

Previous studies on the evaluation of aqueducts can be broadly categorized into two aspects. The first type focuses on studying the influence of various loads on the mechanical properties of hydraulic concrete such as aqueducts. For example, the impacts of seismic loads [15,16,17], wind loads [18], cyclic loads [19] and explosive loads [20] on the performance and failure modes of aqueducts have been analyzed through experimental and numerical methods. The second type aims to evaluate the performance of aqueducts by updating detection methods. A variety of new concepts and methods for detecting aqueducts have been put forward by researchers. For example, the nonlinear dynamic time history of aqueduct structures affected by earthquakes was proposed based on the Dynamic Analysis Method (DAM) [21]. Through the analysis of the dynamic performance and collapse process, the complete instability and failure of aqueducts could be predicted by using effective energy [22]. Additionally, Acoustic Emission (AE) technology has been employed for damage monitoring on hydraulic concrete [23]. In recent years, with the support of artificial intelligence, the dynamic testing and the construction of the Back Propagation (BP) neural network have also been applied to the evaluation of aged aqueducts [24]. However, there are few relative studies that take the concrete as the research object to evaluate the aged aqueducts. The changes in concrete microstructure and the mechanisms remain unclear, and maintenance recommendations for aged aqueducts need to be proposed. For the performance evaluation of aged aqueducts, it is crucial to analyze the mechanical properties and microstructure of concrete.

Among all the deterioration of aqueducts, carbonation of concrete accounts for a high proportion. The CO₂ gas enters the interior of the concrete and reacts with alkaline hydration products, resulting in a decrease in the pH value of the concrete and changes in the pore structure, which reduces the mechanical properties of the concrete [25,26,27,28,29]. In addition, the acidic environment brought by carbonation will also cause corrosion and expansion of rebars. Under the effect of expansion stress, more micro-cracks will be generated inside the concrete, and eventually the mechanical properties of the structure will deteriorate seriously [30,31,32]. The changes in the pore structure and the development of micro-cracks will also reduce the impermeability of the aqueduct [33], which is obviously not conducive to the performance of the water conveyance function. In order to evaluate the performance of the aqueduct more accurately, it is essential to investigate the concrete microstructure such as carbonation and pore structure.

Scanning electron microscopy (SEM) and mercury intrusion porosimetry (MIP) are common and effective methods for microstructure characterizing of concrete. Therefore, the rebound method, the SEM scan and MIP test were taken in this study to investigate the changes of compressive strength and internal microstructure in concrete after 56 years of service. In addition, based on the carbonation and pore structures of concrete, the performance evaluation models for aged aqueducts were proposed, laying a theoretical foundation for practical engineering applications.

2. Experimental Program

2.1. Rebound Method

As a protected structure, standard compressive samples of the aqueduct were not allowed to be cut for safety reasons. To assess the mechanical properties of different parts of the aqueduct after 56 years of service, the rebound method was used to test the compressive strength of the bent columns, beams, arch rings and tank concrete, as shown in Figure 1. These four components encompass the water conveyance and load-bearing structures of the aqueduct, and their compressive strength can sufficiently reflect the concrete quality of its key sections. The aqueduct studied in this paper has total length of 331.3 m, with a design flow of 2.86 m³/s. And there were 42 bent columns, 172 beams, 4 arch rings and 16 tanks in this aqueduct.

The type of rebound hammer was HT225W manufactured by Beijing Koncrete (Beijing, China), with a detection range of 10–60 MPa. The testing process followed the specifications of the Technical Specification for Inspecting of Concrete Compressive Strength by Rebound Method (JGJ/T23-2011) [34]. Each structural component was divided into 10 measurement zones to overcome the heterogeneity of concrete. And in each measurement zone, 16 rebound values were recorded. The highest and lowest 3 values were discarded, and the rebound value of the measurement zone was calculated as the average of the remaining 10 values [34].

The detection direction is horizontal, so additional angle correction is not needed. Finally, the presumed compressive strength of the measurement zones could be obtained by combining the carbonation depth. The average presumed compressive strength of 10 measurement zones was taken as the compressive strength of each component.

2.2. Samples Preparation

The tank of the aqueduct is the most critical structure for water conveyance. During service, the bottom section of the tank remains almost continuously submerged, while the side section experiences dry–wet cycles due to the water flow variations. To further investigate the internal micropore structure of the 56-year-old aqueduct concrete and the influence of water flow on microstructural changes, two cores were cut to obtain samples for SEM scans and MIP tests. One core was taken from the side of tank, and the other was from the bottom of the tank. The diameter of the collected cores was 100 mm. SEM scans and MIP tests were conducted on both the external (close to water) and internal parts of the concrete cores from these two locations.

According to the inspection report [35], the maximum carbonation depth of the aqueduct concrete is 25–30 mm. Therefore, it is expected that the greatest change in the concrete microstructure will occur within the range of 0–30 mm. To study the microstructure changes of the concrete within this range, samples were cut from the external 0–10 mm and the internal 20–30 mm of the taken concrete, with the sampling positions as shown in Figure 2. To obtain more accurate results, two samples were taken from each part, and the sample size was less than 10 mm × 10 mm × 10 mm. The dimensions of specimens were determined according to the standard of Testing and Analysis Methods for Cement-based Materials [36].

The aqueduct had been in service for over half a century, and its concrete hydration process was considered complete. Thus, hydration termination treatment was not performed in this study. Following cutting, the samples were dried in vacuum drying oven at 60 °C for three days. After cooling, all specimens were sealed in vacuum-packed bags and stored until testing commenced. Detailed specimen information is provided in

Table 1. Samples table.

Samples	Sampling Position	Sampling Depth	Qty. of Samples
S-E	Side-External	0–10 mm	2
S-I	Side-Internal	20–30 mm	2
B-E	Bottom-External	0–10 mm	2
B-I	Bottom-Internal	20–30 mm	2

, and the process of specimen preparation is illustrated in Figure 3.

2.3. Microstructural Analysis

2.3.1. SEM Scan

In order to study the microstructure changes of the old concrete of the aqueduct which has been in service for 56 years under the hygrothermal environment, the microstructure of samples S-E and S-I were observed by SEM. The type of the SEM instrument was JSM-IT500 (JEOL Ltd., Tokyo, and Japan), and the scanning was conducted in a high-vacuum environment. Before the experiment, the samples were fixed on the sample stage with conductive glue, and a layer of metal film was uniformly plated on the surface of the samples to enhance the conductivity and prevent the accumulation of surface charges from affecting the images [35]. The plated samples were installed on the center of the scanning stage to ensure that the observation area was located at the center of the field of view. The experimental process is shown in Figure 4. The accelerating voltage of the machine was set as 20 kV, with a working distance of 10 mm.

2.3.2. Mercury Intrusion Porosimetry (MIP) Test

To investigate pore structure changes in the external and internal concrete of the old aqueduct, MIP tests were conducted on four sample groups to obtain their porosity, pore size and pore size distribution. The principle of MIP test can be expressed as Washburn equation, as shown in Equation (1) [37]. MIP tests were carried out using an AutoPore V9600 (Micromeritics, Norcross, GA, USA) high-performance fully automatic mercury intrusion porosimeter. The maximum absolute pressure was set at 61,000 psia (420.6 MPa), and the contact angle was set as 130°.

$$P=-{\displaystyle \frac{2\gamma \cos \theta }{r}}$$

(1)

where P is the pressure applied in the MIP test (Pa); γ is the surface tension of mercury (take 0.48 N/m); θ is the contact angle between mercury and the sample (in this test, it is 130°); r is the pore radius (m).

3. Results and Discussion

3.1. Compressive Strength

The average compressive strength of concrete in four structural components for the aqueduct (bent columns, beams, arch rings, and tank) was assessed via the rebound method. The presumed compressive strength was derived by combining rebound values with carbonation depth measurements, as summarized in

Table 2. Presumed compressive strength by rebound method.

Testing Position	Strength Grade	Testing Compressive Strength (MPa)	Carbonation Depth (mm)	Presumed Compressive Strength (MPa)
Bent column	C18	43.9 (1.87)	27.00 (3.87)	30.0
Beam	C18	42.6 (0.92)	24.50 (2.06)	28.3
Arch ring	C23	41.1 (1.51)	4.00 (0.20)	31.6
Tank	C18	47.2 (2.85)	3.00 (0.54)	44.1

Note: Values in parentheses indicate the standard deviation.

. Carbonation depth data were sourced from the inspection report [35], and the conversion of rebound values to compressive strength followed the Technical Specification for Inspecting of Concrete Compressive Strength by Rebound Method (JGJ/T23-2011) [34].

In this study, the grade of concrete in the main parts of the aqueduct are R200 and R250, which correspond to C18 and C23 in the current specifications, respectively. Unfortunately, the original compressive strength and porosity of concrete were not well documented and preserved due to excessive passage of time. Researchers have proposed probabilistic analysis and a time variation model of concrete compressive strength in existing buildings [38,39]. Based on their research, the original compressive strength can be presumed as 25.1 MPa (C18) and 30.7 MPa (C23), respectively.

Rebound test results indicated that the compressive strength of concrete in the frame columns, beams, arch rings, and tank exceeded the original compressive strength by 19.4%, 12.6%, 3.1%, and 75.6%, respectively. It demonstrates that even after 56 years of service, the concrete of this aqueduct retains sufficient compressive strength to meet operational requirements. Although construction techniques were relatively primitive during the original build phase, stringent material selection controls and rigorous quality supervision ensured structural integrity. Combined with appropriate maintenance practices during its service life, these measures have effectively mitigated further deterioration while sustaining high compressive strength. However, the compressive strength of other parts was higher than the arch ring concrete. This discrepancy may stem from construction factors. Although the arch ring concrete has a higher designed strength, it is located above the river, and the construction conditions at that time were limited, resulting in the quality of concrete placement there being inferior to that in other areas. These findings suggest that with systematic inspection and maintenance protocols, aging aqueducts in a hygrothermal environment can continue to meet the compressive strength requirement. It is worth noting that compressive strength is just one of the key performance indicators of aqueducts. Other failure modes like durability degradation or structural instability should also be evaluated to ensure safety during service life.

3.2. SEM Images

The SEM microanalysis was conducted on two sample groups collected from the side of the tank. The variations of the microstructure in the external and internal concrete were examined at magnifications of 50×, 100×, 300×, 500×, 1000×, 3000×, and 5000×. Three representative images at 100×, 1000×, and 3000× magnifications were selected for illustration, as shown in Figure 5.

Analysis of the 100× magnification images revealed the presence of pores and micro-cracks in both external and internal concrete surfaces. On the 1000× magnification images, the concrete surfaces were observed to be almost entirely covered by calcium carbonate particles (the morphological features observed in Figure 5 are consistent with the typical microstructure of calcium carbonate reported in previous studies [29,40,41]), with micro-cracks becoming more pronounced. It indicates that severe carbonation of the aqueduct during the 56-years service period is a critical factor affecting structural performance.

The 3000× magnification images demonstrated different carbonation patterns between external and internal concrete. Externally carbonated concrete exhibited smaller and denser calcium carbonate particles, while that of internally carbonated concrete was larger and sparser. Fine calcium carbonate fills the pores of external concrete due to more severe carbonation, thus the internal concrete displayed higher porosity and larger pore diameters compared to the external regions. This differential behavior stems from earlier carbonation initiation in external concrete, where calcium carbonate crystallization created a denser microstructure, effectively reducing CO₂ diffusion rates and permeability, thereby partially inhibiting further internal carbonation [42].

3.3. Porosity and Distribution of Pores

Porosity and pore size distribution are the important factors that influence the performance of concrete. The average porosity and average pore diameter of the four groups of samples is shown in Figure 6. For the concrete in the side section of the tank, the external and internal regions exhibited average porosities of 10.98 ± 0.37% and 12.12 ± 0.67%, with corresponding average pore diameters of 11.16 ± 0.15 nm and 11.78 ± 2.62 nm, respectively. This finding aligns with SEM images where external concrete demonstrated more advanced carbonation, forming denser crystalline structures of calcium carbonate that partially filled interstitial voids, resulting in reduced porosity and smaller pore dimensions compared to internal concrete.

The average porosity of the external and internal concrete at the bottom of the tank was 17.57 ± 0.66% and 12.26 ± 0.68%, respectively, with average pore diameters of 11.99 ± 0.24 nm and 7.86 ± 0.57 nm. Unlike the concrete on the side section of the tank, the external concrete exhibited both higher porosity and larger pore diameters than the internal concrete. This difference arises from water flow, which manifests in the following two aspects: Firstly, water flow causes physical scouring on the surface of the concrete. When the scouring force exceeds the bonding force of the internal particles of the concrete, the fine and tiny particles and crystals filling in the pores are washed away, resulting in an increase in porosity and pore diameter. Secondly, water flow also brought chemical erosion and dissolution effects to the concrete. Although the calcium carbonate crystals formed by carbonation were insoluble in water, the acidic substances in the water further reacted with calcium carbonate to form calcium bicarbonate that could be dissolved in water, which also caused an increase in the porosity and pore diameter of the concrete. Enhanced porosity and pore diameter in the external concrete further facilitated the ingress of water and CO₂, accelerating carbonation. The calcium carbonate crystals formed were severely filled in the pores, resulting in the internal pore diameters of the bottom concrete being much smaller than those in other parts.

In addition to porosity, the mechanical performance of concrete is also significantly influenced by pore size distribution [42,43,44]. Cumulative intrusion is essentially the total volume of all pores that are smaller than or equal to the maximum pore diameter corresponding to a given pressure. The pore size distribution of concrete at four locations are illustrated in Figure 7 and Figure 8. The voids can be classified into three categories: gel pores (<10 nm), capillary pores (<1000 nm), and macropores (>1000 nm) based on pore diameter.

MIP test results showed that both the bottom and side sections of the tank concrete mainly contained gel pores and capillary pores, with only a small number of macropores. The proportion of capillary pores in the two side concrete locations was significantly higher than that in the bottom concrete, indicating that the presence of flowing water at the bottom promoted the transformation from capillary pores to gel pores. It is worth noting that for cement-based materials, although carbonation reduces total porosity, it increases the proportion of gel pores and capillary pores [29]. The compressive strength of concrete is almost not affected by the existence of gel pores and smaller capillary pores [45], which explains why the compressive performance still meets service requirements. Nonetheless, the interconnected pores formed by these small pores can increase concrete permeability [33,46]. Excessive permeability not only hinders the water conveyance capacity of aqueducts but also reduces their durability.

The Design Code for Hydraulic Concrete Structures (SL 191-2008) [47] specifies that, to satisfy the normal service performance of hydraulic concrete such as aqueducts, the impermeability grade of concrete should be no less than P8. Previous studies [48,49] have shown that the porosity of hydraulic concrete should range from 5% to 10%. Thus, the porosity of concrete in this study is overly high, failing to meet the service requirements of the aqueduct. Considering this phenomenon, the impact of carbonation and microstructure change on impermeability can be investigated in a future study. The relevant research results may provide support for the enhancement of water conveyance capacity.

4. Mechanical and Hydraulic Evaluation Model of Aqueduct

Microstructural analysis of aqueduct concrete revealed that carbonation was one of the primary deterioration mechanisms. Although the compressive strength of the concrete remained compliant with operational requirements, excessive carbonation may lead to detrimental consequences including concrete cracking, reinforcement corrosion and the degradation of impermeability. Consequently, performance evaluation of aqueduct structures through systematic assessment of concrete carbonation depth and porosity characteristics represents a scientifically valid and practically essential approach.

4.1. Carbonation–Compressive Strength Assessment Model

The factors influencing concrete carbonation are multifaceted, with material properties, environmental conditions and construction practices all contributing to its progression. Depending on the factors considered, researchers have developed various concrete carbonation models in recent years. It is widely accepted that the carbonation depth is characterized by a parabolic trend with carbonation time [40,50], as shown in Equation (2).

$$x_{\mathrm{c}}=k\cdot t^{1/n}$$

(2)

where x_c is the carbonation depth (mm), k is the carbonation coefficient influenced by multiple factors, and t is the exposure duration (year). For the aged aqueduct in this paper, n = 2 can be used because it was built with normal concrete [51].

As shown in Equation (2), the value of k is a key factor affecting carbonation. Taking concrete compressive strength as the main variable, researchers proposed an empirical Equation (3) [52]. Since compressive strength is easy to measure and can comprehensively reflect the influence of materials, curing, and construction factors on concrete performance, this model has certain practical significance.

$$x_{\mathrm{c}}=k\cdot \sqrt{t}={\mathit{\alpha }}_1\cdot {\mathit{\alpha }}_2\cdot {\mathit{\alpha }}_3\left({\displaystyle \frac{60.0}{f_{\mathrm{cu},\mathrm{k}}}}-1.0\right)\sqrt{t}$$

(3)

where f_cu,k is the characteristic value of concrete compressive strength (MPa), and α₁, α₂, α₃ represent the correction coefficients for curing conditions, cement type, and environmental conditions, respectively. α₁ can be determined under the guidance of the Standard for Design of Concrete Structure Durability (GB/T 50476-2019) [53]; for ordinary Portland cement, α₂ = 1.0; for Portland slag cement, α₂ = 1.3; for fly ash cement, α₂ = 1.8; for standard environment, α₃ = 1.0; for dry environment, α₃ = 0.8~0.9; for wet environment, α₃ = 1.1~1.2.

For the tank of the aqueduct in this study, the carbonation depth, compressive strength and carbonation time are 3 mm, 44.1 MPa and 56 years, respectively. Substituting these three parameters into Equation (3), A = α₁$·$α₂$·$α₃ = 1.11 can be obtained. This result reaffirms that the carbonation of concrete has not significantly compromised the compressive strength of the aqueduct when construction quality and long-term maintenance are properly ensured.

4.2. Porosity–Compressive Strength Assessment Model

Porosity and pore distribution constitute critical characteristics of concrete microstructure, exerting significant impacts on both mechanical properties and durability performance. In recent years, porosity–compressive strength relationships have been extensively established through mathematical modeling frameworks [54,55,56].

$$\mathrm{Balshin}\mathrm{model}{\sigma }_{\mathrm{c}}={\sigma }_{\mathrm{c},0}{\left(1-P\right)}^{\mathrm{n}}$$

(4)

$$\mathrm{Ryshkevitch}\mathrm{model}{\sigma }_{\mathrm{c}}={\sigma }_{\mathrm{c},0}{\mathrm{e}}^{-k,P}$$

(5)

$$\mathrm{Schiller}\mathrm{model}{\sigma }_{\mathrm{c}}=K_{\mathrm{s}}\ln \left(P^{\ast }/P\right)$$

(6)

where σ_c is the compressive strength of concrete, σ_c,0 is the compressive strength at zero porosity, P is the porosity of concrete, n, k, K_s and P^* are the empirical constant.

However, the mentioned three classical models above neglect the stress concentration induced by pores, thus incurring inherent deviations. To address this limitation, scholars [55] have integrated brittle fracture theory with the influence of porosity on fracture toughness, developing an advanced porosity–compressive strength relationship model that explicitly incorporates both pore-induced stress concentration effects and stochastic pore distribution characteristics:

$${\sigma }_{\mathrm{c}}={\sigma }_{\mathrm{c},0}{\left[{\left({\displaystyle \frac{P_{\mathrm{c}}-P}{P_{\mathrm{c}}}}\right)}^{1.85}\cdot \left(1-P^{2/3}\right)\right]}^{1/2}$$

(7)

where P_c is the percolation porosity at failure threshold.

In the above strength–porosity equations, the compressive strength correlates with the strength at zero porosity (σ_c,0). Notably, this parameter is not measurable in the laboratory and is treated as an empirical constant. While not precisely obtainable, researchers have proposed a wide range for values of different concretes [57], which can provide theoretical support for engineering applications.

4.3. Porosity–Impermeability Assessment Model

4.3.1. Effect of Aqueduct Pore Structure on Permeation Rate

Equation (8) represents an enhanced permeation rate expression derived from the modified Darcy’s law [58]. As a porous material, concrete subjected to water immersion experiences two dominant permeation driving forces: hydraulic head pressure and capillary pore pressure, while the permeation resistance originates from capillary drag. Both capillary pore pressure and drag are intrinsically related to the size distribution of capillary pores (10–1000 nm) within concrete. Although reduced capillary pore diameters enhance capillary drag (improving impermeability), they simultaneously amplify capillary pore pressure (increasing permeation capacity). The relative influence of these opposing forces on concrete impermeability varies with penetration depth and structural thickness.

$$v=KA\left(\Delta h+p-h_{\mathrm{f}}\right)/L$$

(8)

where v is the water permeation rate (m/s), K is the permeability coefficient, A represents the cross-sectional area of the concrete (m²), Δh is the hydraulic head difference acting on the concrete (m), p is capillary pore pressure (Pa), h_f indicates the capillary resistance (Pa·s/m), L defines the penetration depth (m).

Theoretically, capillary drag increases proportionally with penetration depth, whereas capillary pore pressure remains depth-independent. The critical penetration depth L₀ is defined as the depth at which capillary drag equals capillary pore pressure. When the actual penetration depth L is below L₀, the permeation rate increases inversely with the capillary pore diameter. MIP results reveal that aqueduct concrete pores predominantly fall below 100 nm, with carbonation further reducing pore sizes. This pore refinement paradoxically elevates water permeation rates, adversely affecting hydraulic transport efficiency.

4.3.2. Effect of Aqueduct Pore Structure on Penetration Depth

The frictional resistance of fluid flow in porous material conduits can be determined via Equation (9).

$$H_{\mathrm{f}}={\displaystyle \frac{\lambda L{v}^2}{2dg}}$$

(9)

where H_f is the frictional resistance (N); λ is the Darcy friction factor; L is the penetration depth (m); v is the penetration rate (m/s); d is the diameter of pore (m); g is the gravitational acceleration (m/s²).

When liquids permeate through internal pores of concrete, the frictional resistance is equivalent to capillary drag. According to the Cantor equation, this capillary resistance can be calculated using Equation (10):

$$H_{\mathrm{f}}=h_{\mathrm{f}}={\displaystyle \frac{2\gamma \cos \theta }{r}}$$

(10)

where γ is the surface tension of liquid (N/m); θ is the liquid–solid contact angle (°); r is the pore radius (m).

Substituting Equations (8) and (10) into Equation (9) yields the following expression:

$$H_{\mathrm{f}}={\displaystyle \frac{\lambda L{v}^2}{2dg}}={\displaystyle \frac{\lambda L{\left(KA\left(\Delta h+p-h_{\mathrm{f}}\right)/L\right)}^2}{2\cdot \left(2r\right)g}}=h_{\mathrm{f}}={\displaystyle \frac{2\gamma \cos \theta }{r}}$$

(11)

By rearranging Equation (11), when the capillary resistance balances capillary pore pressure (p = h_f), the critical penetration depth L₀ of water in concrete can be formulated as follows:

$$L_0={\displaystyle \frac{\lambda K^2{A}^2{\left(\Delta h\right)}^2}{2g\gamma \cos \theta }}$$

(12)

Equation (12) reveals that the critical penetration depth of water in concrete shows no direct correlation with pore size. However, it is worth noting that although carbonation decreases the total porosity of concrete, it leads to an increase in interconnected porosity at the microscopic level [30]. Furthermore, carbonation also induces the propagation of micro-cracks, and the flowing water exacerbates the calcium leaching of concrete. All these factors lead to an elevated permeability coefficient, thereby accelerating impermeability degradation in concrete structures.

These above equations illustrate the relationships between carbonation, compressive strength, porosity and permeation parameters of aqueduct concrete, respectively. In engineering applications, the required data can be inferred qualitatively based on the specific index, which will provide a theoretical support for the evaluation of the mechanical and impermeability performance of the aqueducts. However, due to the limited available data and similar research, the comprehensive equations considering the interconnection of these variables still need further investigation in the near future.

5. Conclusions

An old aqueduct in Hunan Province which has been in service for 56 years was investigated in this study. The compressive strength of concrete in different parts was measured by the rebound method, and the microstructural changes were analyzed by SEM scans and MIP tests. The main conclusions are as follows:

1.

The compressive strengths of concrete at the bent columns, beams, aqueduct tank, and arch rings were 30.0 MPa, 28.3 MPa, 31.6 MPa, and 44.1 MPa, respectively. The results exceeded the designed strength by 31–75.6%, indicating that the structure still meets the compressive strength requirements.

2.

Carbonation is a critical factor affecting the performance of the aqueduct. For the side section of the aqueduct tank, the average porosity of the external and internal concrete was 10.98 ± 0.37% and 12.12 ± 0.67%, with average pore sizes of 11.16 ± 0.15 nm and 11.78 ± 2.62 nm, respectively. Consistent with the SEM results, the external concrete exhibited smaller and denser pores. This is attributed to severe carbonation on the exterior, where dense calcium carbonate crystals formed, inhibiting further internal carbonation. Therefore, carbonation resistance and the changes in concrete microstructure should be given priority attention for the monitoring of aqueducts.

3.

The microstructure of the aqueduct concrete was affected by flowing water. For the bottom section of the aqueduct tank, the average porosity of external and internal concrete was 17.57 ± 0.66% and 12.26 ± 0.68%, with average pore sizes of 11.99 ± 0.24 nm and 7.86 ± 0.57 nm, respectively. Physical scouring and chemical erosion caused by flowing water increased porosity and enlarged pore sizes of concrete adjacent to water, with the porosity exceeding the recommended level.

4.

After 56 years of service, the pores in the concrete were primarily gel pores and capillary pores, with sizes below 1000 nm. Although these kind of pores do not significantly reduce compressive strength, it has adverse effects on the conveyance capacity of the aqueduct. For the maintenance of aged aqueducts, not only mechanical properties but also impermeability should be enhanced.

5.

Finally, performance evaluation models for the aqueduct were proposed based on three aspects: carbonation–compressive strength, porosity–compressive strength, and porosity–impermeability relationships, respectively. These models provide theoretical support for the performance evaluation of aged aqueducts. Nevertheless, further investigation and validation are needed in the near future due to the limited data available.