Research on Corrosion Damage and Bearing Characteristics of Bridge Pile Foundation Concrete under a Dry-Wet-Freeze-Thaw Cycle

This study investigated the corrosion damage and bearing characteristics of bridge pile foundations under a dry-wet-freeze-thaw cycle of composite salt in an alpine salt marsh area using an in situ test, laboratory test, and numerical simulations. The in situ test showed that the dry-wet-freeze-thaw cycle has little eﬀect on the quality of the concrete specimens and rebar. The area of the rebar at a depth of 0.25m had the highest corrosion rate of 91%. The application of epoxy resin on the surface improved erosion resistance. After one year of outdoor dry-wet-freeze-thaw cycle test, due to the interaction of compound salts, the quality of specimens is reduced under the denudation of chloride ions, while the mass of specimens is increased by the corrosion products formed under the joint action of carbonate and sulfate, resulting in less obvious change of specimen quality, the antierosion coeﬃcient of the specimens decreased, the maximum loss rate of compressive strength was 38.2%, and the pile foundation began to deteriorate. The laboratory test showed that expansive substances, such as Friedel salt, appeared in the concrete specimens of pile foundation during 225 cycles of dry-wet-freeze-thaw cycles, the relative dynamic elastic modulus was reduced by 60.9%, the antierosion coeﬃcient was reduced to 0.51, and the compressive strength loss rate was 65.9%. As such, the pile foundation was seriously damaged. The numerical simulation shows that, with an increase of the peeling thickness and the corrosion depth, the bearing capacity of the pile foundation will gradually decrease after 8 years. Without maintenance, the bearing capacity of a pile foundation will decrease by 34.45% in the 20 th year.


Introduction
With the promotion of the Western Development Strategy in China, "development driven by traffic" has become an important part of construction in Qinghai Province. However, terrain conditions are complicated. During the construction of roads and bridges, many problems are encountered. Among them, corrosion damage and bearing characteristics of pile foundations is a concern [1][2][3][4][5]. Feng et al. [1][2][3] analyzed the relative dynamic elastic modulus and compressive strength of pile foundation concrete in an in situ test and predicted the life of the pile foundation without maintenance. Murad et al. [6,7] used the in situ tests to study the bearing characteristics of a pile foundation of the I-10 double-span bridge in Louisiana. According to the fitting curve and mathematical derivation, it was considered that there was no group effect in the derived p-y curve. Pilehvar et al. [8][9][10][11] studied the mechanical properties and durability of pile foundation concrete after replacing cement with fly ash under the conditions of freeze-thaw cycles in laboratory model tests. Ghazy et al. [12][13][14] studied the mechanical properties and durability of concrete specimens under the conditions of indoor freeze-thaw cycles, as well as the microscopic changes of specimens using scanning electron microscopy.
Other researchers have focused on pile foundations under special circumstances. Shi et al. [15,16] studied the axial loadbearing characteristics of cone-pillar pile foundations in noncohesive soil under freeze-thaw cycles by laboratory tests and proposed an appropriate design scheme. Feng et al. [1,17] studied the vertical load-bearing characteristics of concretefilled steel, tube-reinforced concrete piles in a loess area based on a centrifugal model test. Dong et al. [18] discussed liquefaction identification of saturated silty sand of bridge pile foundations under simulated earthquake conditions using a large-scale shaking table test of a laminated shear model box. Ali et al. [19,20] studied the bearing characteristics of bridge pile foundations in expansive soil, proposing a corresponding probability model; they verified the feasibility of the models based on practical cases. Yan et al. [21] studied the mechanical properties and deformation characteristics of pile group foundations nested in inclined weak interlayers with a centrifugal model test. Feng et al. [22] studied the distance effect and bearing characteristics of pile foundation under faultpile-rock-soil interaction.
Some researchers have used numerical simulation to study the mechanical properties of pile foundations. Yao et al. [23,24] studied the vertical bearing characteristics of bridge pile foundations in salt marshes with different spalling thicknesses and corrosion depths by manipulating the pile diameter and pile length in numerical simulations. Weyers [25] studied the service life of reinforced concrete in a chlorine-containing environment by numerical simulation and established a corresponding prediction model. Lang et al. [26,27] studied the bearing characteristics of four different types of marine bridge pile foundations by a numerical simulation model. Feng et al. [28] studied the bearing characteristics of bridge pile foundations under the condition of liquefaction of foundation soil by improving a numerical simulation calculation method.
Predecessors used different methods such as the in situ test, laboratory test, and numerical simulations to study the mechanical properties of bridge pile foundations under dry-wet or freeze-thaw cycles, but the research methods and research environmental conditions are relatively simple, and few people use these three methods to study the bearing characteristics and corrosion damage of pile foundation under the dry-wet-freezethaw cycles. In addition, many researchers have studied the mechanical properties of pile foundations under special conditions such as cohesive soil, loess, expansive soil, weak interlayers, and earthquakes, but few have studied the bearing characteristics and corrosion damage of pile foundation in the special environment of alpine salt marsh area.
is paper explores this dynamic in an alpine salt marsh area with the in situ test, laboratory test, and numerical simulations. Based on the results, engineering suggestions are given to provide a reference for similar projects.

Project Overview.
Dexiang Expressway is an important traffic route from Qinghai Province to the hinterland of Haixi Prefecture with a total length of 165 km. Affected by regional climate, salt marshes are widely distributed along the route, as shown in Figures 1(a) and 1(b). Due to the high wind and low rain, as well as the large temperature differences between day and night, it is affected by a strong drywet-freeze-thaw cycle, resulting in many structural problems, as shown in Figure 1. e salt water-influenced section of the main line of the Dexiang Expressway is about 22.7 km. e makeup of the road sections is shown in Figure 2.  Table 1, and location distribution is shown in Figure 3. e burial ages were 3, 9, and 12 months.

Test Index
(1) Quality. e quality of the specimen was weighed using an electronic balance with an accuracy of 0.1 g, which can reflect the quality change of the specimen before and after corrosion.
(2) Antierosion Coefficient. e antierosion coefficient is used to describe the degree of corrosion resistance of the specimen, and it is calculated as follows: where K c is the antierosion coefficient of the specimen; R c is the compressive strength of specimen immersed in erosion solution, MPa; and R s is the compressive strength of specimens immersed in water at the same age, MPa.
(3) Compressive Strength. After the specimens reached the specified ages, they were removed and weighed (the specimens in water were washed and dried before weighing), and the uniaxial compression resistance of the specimens was performed on the electrohydraulic servo universal testing machine (pressure 2000 kN) of North China University of Water Resources and Electric Power, Zhengzhou, China.
(4) Corrosion Rate. e bridge pile foundation needs to be reinforced during the construction process. To simulate an actual situation, three kinds of rebar were used in the construction process: φ12 mm, φ25 mm threaded rebar, and one with the surface coated with epoxy resin on φ25 mm threaded rebar (φ25′). When preembedded, the rebar was first allowed to rust. After reaching the specified ages, they were removed; see Figure 4. After the rebars were taken out, the area corrosion rate and quality corrosion rate were measured. Before measurement, a full cleaning was carried out to remove the   concrete on the surface of the rebar. e calculation methods are given as equations (2) and (3): Area corrosion rate: Quality corrosion rate: where S 0 is the surface area excluding both ends of the rebar; S 1 is the corrosion area of the rebar surface; and m 0 and m 1 are the masses before and after the corrosion of the rebar. When calculating area corrosion, put the rebar on the graph paper and draw the corrosion area, so as to calculate the corrosion area of the rebar. e area corrosion rate calculation is shown in Figure 5.

Laboratory Test.
e in situ test could not be used to simulate the long-term performance of the pile foundations after corrosion. erefore, indoor accelerated tests were used to examine mechanical properties and microscopic mechanisms of pile foundations with corrosion.

Materials.
e test employed different materials (carbonate, sulfate, and chloride) and climates (dry-wet and freeze-thaw cycle) to simulate the actual working conditions of the bridge pile foundation concrete under salt dry-wetfreeze-thaw cycles. According to the onsite investigation, the content of each ion in the compound salt solution is shown in Tables 2 and 3.
Indoor and in situ tests used the same proportion of concrete specimens. ey were divided into two types: one was a block with side lengths of 100 mm used to measure the mass and relative dynamic elastic modulus of the specimen before and after the test. e other was a prism (100 mm × 100 mm × 400 mm) used to measure the compressive strength and microstructure of the specimens. To better simulate the service environment of the pile foundation concrete, the formed specimens were left for 24 hours, the molds were removed, they were allowed to cure to the specified ages, and then the tests were carried out.

Test Procedure.
After curing the specimens for 24 days, they were immersed in the compound salt solution prepared in advance (4 days prior). After removing the specimens and freezing them at −15 ± 2°C for 2 hours, they were thawed at 6 ± 2°C for 2 hours. is operation was a freeze-thaw cycle: samples were placed in an oven at 80 ± 5°C for 8 hours, then cooled for 1 hour, and soaked for 15 hours.
is operation was a dry-wet cycle [2,3,23]. e quality and dynamic elastic modulus of the specimens were measured 25 times. Representative specimens were selected for the compressive strength test and scanning electron microscope (SEM) test and the chemical elements were analyzed by EDS as shown in Figure 6. ree sets of specimens are selected for research each time.

Test Index
(1) Quality loss rate: where ΔW n is the quality loss rate of the specimen; G 0 is the initial mass of the specimen, g; and G n is quality of specimen after a certain number of cycles, g. (2) Relative dynamic elastic modulus: where E r is relative dynamic elastic modulus; f 0 is the initial frequency of the specimen measured with a dynamic elastic modulus tester; and f n is the frequency of specimen after N test cycles. (3) Compressive strength: In order to make the test results more accurate, three specimens were measured by an electrohydraulic       [23,24,29], the peeling thickness and corrosion depth reflect the severity of the corrosion of a bridge pile foundation. According to the surveys of corrosion of the substructure of bridge pile foundations in the Qinghai area, the corrosion depth is concentrated at 8 m below the ground surface and the peeling rate of the pile concrete is about 3 cm/4 years. So, the reduction in pile diameter is set to 3 cm per 4 years, and the reduction in corrosion depth is set to 1.6 m per 4 years in the model. e peeling thickness and corrosion depth corresponding to different times are shown in Table 4.

Model Establishment.
Marc finite element software was used to simulate the bearing characteristics and corrosion damage of bridge pile foundations for alpine salt marsh regions. e pile diameter was set as 1.8 m, the pile length as 40 m, the model boundary as 8 times the pile diameter [30], and the pile subsoil thickness as 20 m. e geometric model schematic is shown in Figure 7. e foundation soil is regarded as an ideal elastoplastic model, and the spalled area and pile foundation concrete materials are as well. e Mohr-Coulomb yield criterion can be used to better simulate the interaction between piles and soil. In the modeling process, it was assumed [31] that each rock and soil body were homogeneous, isotropic, and continuous.

Parameter Selection.
Wu et al. [32] converted the number of onsite dry-wet-freeze-thaw cycles to the number of indoor dry-wet-freeze-thaw cycles and believed that Qinghai Province would experience an average of 110 dry-wet-freezethaw cycles per year, equivalent to about 11 indoor dry-wetfreeze-thaw cycles. According to the laboratory test, the relative dynamic elastic modulus of the specimens after 225 cycles was reduced by 60.9% at which time it is considered that the pile foundation function fails [33]. In the numerical simulation calculation, the corrosion depth of the pile foundation is simulated based on the attenuation amplitude of the relative dynamic elastic modulus as measured by the dry-wet-freeze-thaw cycle in the laboratory test, and the peeling thickness of the pile foundation is simulated by changing the pile diameter. e model parameters corresponding to different peeling thicknesses and corrosion depths are shown in Table 5.

Method
Analysis. When the model was established, the pile diameter was set to 1.8 m, and the pile length was 40 m for the first time. e width of the outer grid of the pile foundation is set to 3 cm for each layer, the 5 layers are 15 cm in total, and the remaining internal grids are equally divided. Eight metres down from the pile-soil boundary, set as a corrosion depth zone, the grid is divided into 5 parts, each part is 1.6 m high, and the rest of the internal grids are equally divided. Each time the model is calculated, the study of the thickness of the pile foundation spalling is achieved by changing the parameters of the outer grid of the pile diameter from the outside to the inside (each change is 3 cm); the change of the corrosion depth of the pile foundation is realized by changing the length of the corrosion zone from top to bottom (each change is 1.6 m); and after each change is completed, the P-S load-settlement curve is extracted, and the load corresponding to a settlement of 40 mm is the corresponding bearing capacity of the pile foundation [24,28]. e model calculation diagram is shown in Figure 8.

Quality Changes.
e quality changes of concrete specimens with different depths of pile foundations are shown in Figure 9. With an increase in the corrosion time, the quality change increased or decreased under the four embedding conditions, but the overall variation did not exceed 2%. e quality of the specimens in the chloride ion ablation was reduced, and the corrosion products generated under the combined action of carbonate and sulfate increased the quality of the specimens, resulting in an insignificant change in the quality of the specimens under the interaction of the compound salts. e main equation is as follows:

Antierosion Coefficient.
e change of the antierosion coefficient of concrete specimens with different depths of pile foundation is shown in Figure 10, and the regression equations and correlation coefficients corresponding to the antierosion coefficients are shown in Table 6. From the water to 1.25 m underground, the correlation coefficient of each regression equation increases. e antierosion coefficient of the pile concrete specimens at the corresponding time was obtained, which provides a basis for evaluating the durability of the specimen. With the increase in time, the antierosion coefficient at 0.25 m in water, surface, and underground increases and then decreases, indicating that the performance of specimens has begun to deteriorate. With less Advances in Civil Engineering depth, the hydration of earlier specimens is slow and the antierosion coefficient increases over time.

Compressive Strength Loss
Rate. e change of compressive strength loss rate of concrete specimens with different depths of pile foundations is shown in Figure 11.
With the increase of age, the loss rate increased with a maximum loss of 38.2%. For a given age, as the depth increases, the loss rate increases first, then decreases, and finally increases.
is is because the one specimen is immersed in water, and the compressive strength of the specimens underground is higher due to the slow hydration.

Corrosion Rate of Rebar
(1) Area Corrosion Rate. See Table 7 for the corrosion rate of different types of rebar at different depths. Taking φ25 mm as an example, the corrosion rate from water to 1.25 m underground is 76%, 91%, 66%, and 65%. e corrosion rate of the rebar area increases first and then decreases from surface to depth. It can be seen from Table 7 that, in the compound salt environment, the φ25′ rebar coated with epoxy resin exhibits a good rust inhibition. After 360 days of testing, the surface of the rebar has no corrosion, while the rebar without epoxy resin has large corrosion areas as shown in Figure 12.
(2) Quality Corrosion Rate. e rebars with the largest areas of corrosion were selected to measure the quality corrosion rate (Table 8). Although the corrosion rate of the rebar area was large, the mass loss was small, and the mass loss rate is less than 0.5%, which has little effect on the mechanical properties of the rebar. According to in situ test analysis, after one year of drywet-freeze-thaw cycles, the overall performance of the concrete specimens has relatively low changes. e quality of the reinforced concrete specimens is relatively low, but the surface corrosion rate of the steel bars, the compressive strength of the concrete, and the corrosion resistance coefficient have all decreased to varying degrees, indicating that the specimens have begun to show damage.

Appearance.
e failure characteristics of a concrete specimen after 225 cycles of dry-wet-freeze-thaw cycles of composite salt are shown in Figure 13. It can be seen that the concrete at the corners degrades and small cracks appear around this area.

Quality.
e change of mass loss rate of the pile foundation concrete specimens after 225 dry-wet-freezethaw cycles of the composite salts is shown in Figure 14(a). As the number of cycles increased, the mass loss rate of the specimen increased first and then stabilized. At 50 cycles, the Quality (g) Figure 9: Quality change of pile foundation concrete specimens. Compressive strength loss rate (%) Figure 11: Change in the compressive strength loss rate of pile foundation concrete specimens. 8 Advances in Civil Engineering mass loss rate was 1.04%. After 225 cycles, the mass loss was 1.93%, and the overall mass loss change was small.

Relative Dynamic Elastic
Modulus. e sulfate and carbonate products produced by early corrosion increased the relative dynamic elastic modulus of the specimen (Figure 14(b)). As the number of cycles increased, the relative dynamic elastic modulus of the specimen decreased. After 225 cycles, the relative dynamic elastic modulus of the specimen decreased to 60.9%, the specimen tended to break, and the pile foundation was considered insufficient.

Antierosion Coefficient.
As the number of cycles increased, the antierosion coefficient of the specimen gradually decreased (Figure 14(c)). When the cycle was 0-50 times and 75-125 times, the decrease was 0.18 and 0.2. At 50-75 times, the decrease was 0.01. e ettringite or calcite products generated by the corrosion at the beginning of the cycle made the soil around the pile body dense, and the change of the antierosion coefficient of the specimen slowed down. With the increase in the number of cycles, the tensile stress caused by the volume expansion of the corrosion products in the early stage was greater than the tensile stress of the specimen itself, and the cracks in the corrosion area of the specimen generated cracks, which reduced the strength. After 225 cycles, the antierosion coefficient decreased to 0.51, and the damage was most extensive.

Compressive Strength.
e compressive strength loss ratio of the specimen decreased, but the decline was small at first (Figure 14(d)). With the deterioration of specimen performance, the loss rate of compressive strength increased.    Figure 13: Failure characteristics of a concrete specimen.

Advances in Civil Engineering
After 225 cycles, the loss rate of compressive strength was 65.9% and the damage was extensive.

Microscopic Mechanisms.
A microscopic SEM image of the pile foundation concrete specimen after 225 cycles of dry-wet-freeze-thaw cycles with composite salt is shown in Figure 15. e EDS energy spectrum analysis is shown in Table 9. After 225 cycles, the specimen is loose, the amount of gelatinous substance is relatively low, and there is a large gap internally. In the EDS energy spectrum analysis, C, O, Si, Al, Ca, and Cl were detected. is  suggests that after a long period of the dry-wet-freezethaw cycle, the strong penetrating Cl − enters the inside of the specimen. e specimen was damaged, but the intrusion of Cl − also inhibits the entry of SO 4 2− , resulting in a slower erosion. Since the presence of S was not detected in the energy spectrum analysis, it indicated that the rodknitted fabric (yellow-framed) was not ettringite (3CaO·Al 2 O 3 ·3CaSO 4 ·32H 2 O), but possibly Friedel salt (3CaO·Al 2 O 3 ·CaCl·10H 2 O).
Limited by the in situ test conditions, the laboratory accelerated test is used to study the damage of concrete. e research results show that after 225 times of dry-wet-freezethaw cycles, the concrete specimens tend to be damaged, and the relative dynamic elastic modulus, corrosion resistance coefficient, and compressive strength of the specimens all show large damage changes. ere is cl in the specimen, and Friedel salt may appear in the specimen. e appearance of concrete specimens showed partial peeling, and the quality of specimens changed relatively low.

Comparative Analysis of Indoor and In Situ Tests.
After 1 year of in situ tests and 225 cycles of indoor dry-wetfreeze-thaw, the quality change of the pile foundation concrete specimen was shown to be similar, both less than 2%.
is indicated that the dry-wet-freeze-thaw cycle has little impact on the quality changes of specimens. After one year of in situ testing, the antierosion coefficient of pile foundation concrete specimens began to decrease, and the performance of the specimens began to deteriorate. e resulting regression equation had a good predictive ability of the antierosion coefficients. However, it is more difficult to predict the anticorrosion coefficient of specimens after many years, and the laboratory test was used to this end. After one year of in situ testing, the compressive strength loss rate of the concrete specimens of the pile foundation is similar to the results obtained from the laboratory test. With the increase of the number of dry-wet-freeze-thaw cycles and the interaction of various ions, the loss rate of compressive strength of the specimen obtained by the laboratory test generally increases.

Numerical Simulation Results.
e change in the bearing capacity of the pile foundation in different years under different peeling thicknesses and corrosion depths is shown in Figure 16. e initial bearing capacity of the pile foundation is 587.51 kN. With the increase of the peeling thickness and the corrosion depth, the bearing capacity decreases little in the first eight years, and the bearing capacity in the eighth year is 582.92 kN. After eight years, the bearing capacity gradually decreased. In the 20 th year, the bearing capacity is 385.12 kN, a decrease of 202.39 kN which is 34.45%. In the eighth year, the corresponding peeling thickness of the pile foundation was 6 cm and the corrosion depth was 3.2 m. erefore, it is believed that the pile foundation started to show damage around this time.
According to [1,34], the weight of a personal car is 1 ton. e weight of a typical small truck is 5 tons and a large truck is 20 tons. According to previous research, during normal traffic, the ratios of the vehicles are 7 cars : 2 small trucks : 1 large truck. e initial bearing capacity of the pile foundation is 587.51 kN. At this time, it can bear 41 cars, 2 small trucks, and 1 large truck at the same time. After 8 years, the bearing capacity was reduced by 4.59 kN, which is equivalent to 0.5 cars. After 20 years, the bearing capacity is reduced by 202 kN, which is equivalent to 14 cars and a small truck.
Numerical simulation analysis believes that the bearing capacity of the pile foundation will be greatly reduced after 8 years. If no measures are taken, the operation of the pile foundation will be affected. In practice, it is suggested that an appropriate increase of pile diameter or pile length should be combined with long-term pile foundation protection measures to make up for the reduction of pile bearing capacity caused by pile corrosion, so as to ensure the safety of the highway bridge pile foundation during normal operation.

Conclusion
(1) One year after the in situ test dry-wet-freeze-thaw cycle, the quality change of concrete specimen of the pile foundation is not obvious and the maximum loss rate of the compressive strength of it was 38.2%. e change in the antierosion coefficient showed that the performance of the pile foundation began to deteriorate. e corrosion rate of rebar at a depth of 0.25 m was the largest, reaching 91%. When the epoxy resin was applied on the surface of the rebar, the corrosion rate was significantly reduced.
(2) After 225 dry-wet-freeze-thaw cycles (representing about 20 years) in the laboratory test, the corner of the concrete specimen of the pile foundation deteriorated. e quality change of pile foundation concrete specimen is not obvious. e antierosion coefficient gradually decreases to 0.51 and the damage is serious. e relative dynamic elastic modulus was reduced by 39.1%. e compressive strength loss rate was 65.9%. Due to expansive substances, such as Friedel salt, the concrete tended to deteriorate. It is suggested that construction should be completed three months before the freezing period, and unscheduled maintenance and repair be carried out according to the conditions of a given situation.
(3) e bearing capacity of the pile foundation remained unchanged for the first eight years of the wet-dryfreeze-thaw cycle and gradually decreased after eight years, with a decrease of 34.45% in the 20 th year. In later periods of service, attention should be paid to pile foundation protection. In construction, a steel casing can be used to protect the bridge pile foundation. (4) Due to the limitation of in situ test conditions, this paper uses laboratory test and numerical simulation methods on the basis of in situ tests to study the corrosion damage and bearing characteristics of bridge pile foundations under dry-wet-freeze-thaw cycles in the alpine salt marsh area. e similar complex multifactor coupling and special conditions of bridge pile foundation research provide reference and have very important guiding significance.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare no conflicts of interest.