Feasibility Study of Temperature Control Measures during the Construction of Large-Volume Concrete Gravity Dams in Cold Regions: A Case Study

Abstract

Effective temperature control measures are crucial for achieving temperature regulation and preventing cracking in the dam body during the construction of large-volume concrete gravity dams. Due to the low ambient temperatures in winter, it is especially important to focus on temperature control measures for concrete dam construction in cold regions. This paper employs a numerical simulation method that takes into account dam temperature control measures to simulate and predict the overall temperature and stress fields of the Guanmenzuizi Reservoir Dam, and validates these simulations with field monitoring results. This study finds that the ambient temperature significantly impacts the temperature and stress of the dam body’s concrete. The internal temperature of the dam reaches its highest value approximately 7 days after pouring, followed by periodic fluctuations, with the dam body’s temperature changes exhibiting a certain lag compared to the ambient temperature. The interior of the dam is under compression, while the upstream and downstream surfaces experience significant tensile stress. This project adopts targeted temperature control measures for the cold environmental conditions of the region, which are reasonably implemented and effectively reduce the temperature rise of the concrete during construction, achieving the temperature control objectives. This study also explores the impact of the cooling water pipe density on the dam body. The research results offer valuable references for the implementation of temperature control measures and the establishment of temperature control standards for concrete gravity dams in cold regions.

1. Introduction

The harsh climatic conditions of cold regions, such as low environmental temperatures and frequent freeze-thaw cycles, can lead to a decline in the performance of concrete dams in cold areas, posing a threat to the structural safety of these constructions [1]. Permafrost covers approximately one-quarter of the Earth’s terrestrial surface, with cold regions having an even broader distribution. In China, the cold regions are defined by three criteria: the average temperature of the coldest month being below −3.0 °C, the number of months with an average temperature above 10 °C not exceeding five, and an annual average temperature of ≤5 °C; they constitute 43.5 % of China’s land area [2]. Many concrete dams have been constructed in high-latitude cold regions [3]. Over the past decade, China has built more than 87,000 dams, with over 60% located in cold regions [4,5]. Compared to warmer areas, concrete dams in cold regions face challenges, such as large annual temperature variations and lower stable temperatures within the dam body [6,7]. For example, the Ust-Ilimsk concrete gravity dam in Russia has an average annual temperature of just −3.9 °C [8], while the Burqin Shankou double-curvature arch dam in Xinjiang, China, can experience extreme temperatures ranging from 39.4 °C to −41.2 °C, with an annual temperature variation of up to 80 °C [9]. These conditions severely threaten the service performance and structural safety of dams, with the release of hydration heat during construction causing significant temperature differences inside and outside the dam body, potentially leading to temperature-induced cracks [10,11,12,13]. Compared to traditional measures, such as controlling the concrete pouring temperature, surface spraying, and water cooling [14], concrete dams in cold regions usually require additional temperature control measures, such as optimizing the concrete mix design and insulating the dam surface, to reduce the risk of cracking [15].

The characteristics of the internal stress and strain fields within the dam body during construction are crucial for assessing dam safety [16]. Considering the complexity of construction-period issues, accurately simulating the dam temperature and stress are essential, taking into account the construction conditions, boundary conditions, and temperature control measures. Numerical simulation is a vital tool for the ongoing monitoring of large concrete dams [17], requiring a finite element simulation analysis of the entire construction process, and utilizing the on-site pouring data and safety monitoring data to study the trends in the temperature and stress changes within the dam.

Zhang et al. [15] used three-dimensional finite element meshing techniques to simulate the temperature and thermal stress in cold-region roller-compacted concrete gravity dams, comparing the effects of no insulation versus the use of 5 or 8 cm thick polystyrene boards as temperature control measures. The results indicated that surface insulation significantly reduces the maximum tensile stress. Lai et al. [18] proposed a finite element calculation formula for the influence of engineering seepage fields on temperature fields, calculating the impact of seepage fields on the temperature fields of dams in cold regions. Chen et al. [19] introduced an equivalent simulation unit for concrete–polyurethane composite structures and their calculation methods, establishing a finite element model for a concrete arch dam in a northwest cold region, and studying the insulation effects of different thicknesses of polyurethane foam layers. Zhang et al. [6] developed a temperature field calculation program for the construction and operation phases of concrete dams, analyzing the trends in the thermal stress of a Tibet crushed-stone concrete gravity dam with and without insulation measures during construction and operation. Chen et al. [3] proposed a reverse analysis method that couples numerical simulation models with dam safety monitoring data, offering good fitting accuracy and rationality for the thermal parameter inversion of insulated concrete dams in cold regions. Cai et al. [20] used regression analysis to establish a prediction model for the mechanical properties of concrete under freeze–thaw cycles, predicting a precipitous decline in dam structure reliability over long periods. Dam safety monitoring is crucial for understanding the engineering state of dams, requiring a comprehensive analysis of the raw monitoring data to obtain relevant information about a dam’s working state [21,22,23]. Aleksandrov [24] analyzed the impact of the concrete temperature distribution within the dam body on the stress and deformation state of a dam using temperature monitoring data from the Sayano-Shushenskaya hydroelectric power station dam. Some scholars [25,26,27] have researched the finite element simulation reconstruction of concrete dam temperature fields, comparing them with monitoring data to identify the temperature rise patterns during the concrete pouring process.

This paper adopts a numerical simulation method considering the temperature control measures for dams, focusing on a concrete gravity dam under construction in a cold region. It thoroughly considers the unique temperature boundary conditions of cold regions, simulates the temperature control pipe network embedded within the dam body, and establishes a finite element simulation calculation model to ensure that all the parameters and boundary conditions closely match reality. Combined with the actual needs of the project, the simulation analysis and prediction of the overall temperature field and temperature stress of the dam are carried out in stages, and a detailed simulation analysis is carried out on the key parts. By tracking and inverting the thermal and mechanical parameters and boundary conditions of the dam foundation and the dam concrete itself, and combining these with the monitoring data for verification, the distribution and variation in the temperature and stress inside the dam concrete over space and time are dynamically tracked. Suggestions for temperature control measures and temperature stress control standards for concrete gravity dams in cold regions during construction are put forward.

2. Dam Construction and Temperature Control Measures

2.1. Project Overview and Monitoring Instruments

The Guanmenzuizi Reservoir is located in the northeast of Hegang City, Heilongjiang Province, with the dam site situated upstream of the Wutong River (a first-level tributary of the Songhua River). It serves as a hydraulic hub, primarily for the urban water supply and agricultural irrigation, while also providing flood control and facilitating power generation, among other comprehensive uses. The reservoir project is classified as a large Type II scale, with a total capacity of 403 million cubic meters. The river-blocking dam is a normal concrete gravity dam, with a maximum height of 34.38 m and a length of 596 m, divided into 31 dam sections. Both banks consist of water-retaining dam sections, with the spillway dam sections (no. 6, no. 7, and no. 8) located in the riverbed.

The crest elevation of the dam is 150.40 m above sea level, with a normal storage level of 146.50 m and a dead water level of 132.50 m. The design flood level (the recurrence period is 100 years) is set at 147.85 m, and the check flood level (the recurrence period is 1000 years) is set at 150.02 m. The water-retaining sections of the dam use normal concrete with a strength grade of C15, while certain areas (such as the water-facing surface and stress concentration points) use normal concrete with a thickness of 2.5 m and a strength grade of C25. The construction of the project commenced in August 2021, with the pouring of the dam body’s concrete beginning in April 2022. Images of the dam from the upstream and the downstream plan are shown in Figure 1 and Figure 2, respectively. The dam is currently poured to an elevation of 150.40 m. Figure 2 shows that 0 + 088 is the dam pile number, and 0 + 088 is the central section of the no. 5 dam section.

Located in a temperate continental monsoon climate zone, the Guanmenzuizi Reservoir experiences hot, rainy summers, and cold, prolonged winters. According to statistics from the Hegang Meteorological Station, the highest air temperatures occur in July, with an extreme maximum air temperature of 37.7 °C, and the lowest occur in January, with an extreme minimum air temperature of −38.8 °C. The average annual air temperature is 3.3 °C. The average annual sunshine duration is 2443 h, with a frost-free period of 134 days and an average annual freeze-up duration of about 147 days, with a maximum frozen soil depth of 2.0 m.

2.2. Dam Temperature Control Measures

Temperature control during construction remains one of the most critical issues in dam construction. In the design of large-volume concrete, it is commonly assumed that the material primarily withstands compressive stress, with little to no consideration given to the tensile stress [28]. However, in cases of extensive concrete pouring, significant tensile stress may arise if the generated heat is not sufficiently dissipated, leading to temperature fluctuations and potentially causing cracks to exceed permissible limits. Therefore, the careful determination of the concrete composition and the implementation of appropriate cooling schemes to effectively resolve conflicts between temperature stress control and crack prevention are often key to successful dam construction [29].

The average annual temperature in the environment of Guanmenzuizi Dam is low, resulting in the relatively low stable temperature of the dam body, estimated to be only 5–9 °C. However, the highest temperature of the concrete poured in summer can be quite high, with extreme temperatures reaching up to 37.7 °C. The winters at Guanmenzuizi Dam are cold, making it difficult to control the internal and external temperature differences of the upstream and downstream surfaces of the concrete, which can easily cause surface cracks on these faces, potentially developing into deep cracks. Concrete construction on the dam is only carried out from April to October each year.

Given the significant temperature differences at the dam’s foundation and the difficulty in controlling through-cracks in the foundation, it is necessary to keep the pouring temperature in summer to 15 °C, while in other seasons, the concrete can be poured naturally. During high-temperature seasons, measures such as surface spraying and running water on the concrete surface are implemented. The primary cooling of the dam’s concrete is conducted during the initial stages of pouring, using water cooling to control the maximum temperature of the concrete. For each layer of concrete poured between May and August, water pipes are arranged at intervals of 1.5 m × 1.5 m (layer spacing × horizontal spacing) for water cooling, as shown in Figure 3. High-strength polyethylene pipes (with an inner diameter of 32.6 mm and an outer diameter of 40 mm) are used for the cooling water, with a flow rate of about 15–20 L/min. The direction of water flow is alternated daily, and the water temperature is maintained at 15 °C. Water cooling begins 1 day after the concrete is poured and continues for 16 days post-pouring.

The purpose of the secondary cooling is twofold: first, to cool the concrete poured each summer before winter, reducing the internal and external temperature differences in the concrete near the upstream and downstream faces of the dam, and thus reducing the temperature stress during winter. The secondary cooling measures involve using river water for cooling, starting in early November each year, with the cooling water temperature at 10 °C and the target temperature at 22 °C. For the permanently exposed surfaces, such as the upstream and downstream faces, temporary insulation methods are used before October, and permanent insulation methods are adopted after October.

2.3. Dam Safety Monitoring

For this project, over 500 monitoring instruments were installed across 31 dam sections. These instruments primarily perform deformation monitoring, seepage monitoring, stress-strain monitoring, and temperature monitoring. During the construction period, the monitoring data were manually collected on-site. This paper analyzes the temperature monitoring data and stress-strain data for the non-overflow dam sections during the construction period, aiming to investigate the safety behavior of normal concrete gravity dams in extremely cold regions during construction.

2.3.1. Dam Temperature Monitoring

For this project, the thermometers were arranged in a grid pattern on the central profiles of dam sections 5# to 14#. The vertical spacing was 15 m, the horizontal spacing was 8 m, and they were placed 10 cm from the upstream dam face. Additionally, boreholes were drilled upstream and downstream of the dam foundation, with a depth of 5.5 m, and thermometers were embedded at 1.5 m, 3 m, and 5 m below the foundation surface. In the overflow dam sections 6#, 7#, and 8#, 17 thermometers were installed in each section, while in the remaining water-blocking dam sections, 15 thermometers were installed in each section, totaling 156 thermometers. The thermometers used were DW-1 resistance temperature sensors, with a measurement range of −30 °C to +70 °C, a measurement accuracy of ±0.3 °C, and a sensitivity of 0.1 °C. According to the construction schedule, the thermometers were embedded at different elevations within the dam body to monitor temperature changes. This paper analyzes the temperature variation in the dam body during the construction period using section 12# as an example. The positions of the thermometers are shown in Figure 4.

2.3.2. Dam Stress-Strain Monitoring

For this project, the stress-strain monitoring points are arranged at the central profiles of dam sections 0 + 088, 0 + 098, and 0 + 208, with a total of 16 sets of five-direction strain gauge groups. The strain gauges used are NZVS-150E vibrating wire strain gauges, with a measuring range of 3000 με and a measurement accuracy of ±0.1 %FS. For the monitoring data, positive values indicate compression and negative values indicate tension. This paper utilizes the strain gauge data to observe the concrete strain, further analyzes and processes the data, and then applies the relevant mathematical formulas to calculate the stress at the monitoring points. Using section 12# as an example, three sets of five-direction strain gauges (including non-stress gauges), S1, S2, and S3, are installed at an elevation of 121.02 m. The embedded positions and arrangement of the five-direction strain gauges are shown in Figure 5, with the instruments pre-embedded using an installation method.

2.3.3. Outlier Identification of Data

Regarding the dam safety monitoring work during the construction period, monitoring errors in the data collection are inevitable, due to the disturbances caused by the on-site construction. Therefore, it is necessary to detect anomalies in the monitoring data caused by these errors. The monitoring errors are classified into three categories: systematic errors, random errors, and gross errors [30]. For the same monitoring instrument, there may be individual data points in the construction period observation sequence that differ significantly from other data points, causing notable fluctuations in the plotted process line. These anomalous data points may contain gross errors, and need to be identified to determine whether they are outliers within the data set. Appropriate methods must be used to judge whether these anomalous points should be considered outliers [31,32].

The jump characteristic method is a technique used to detect gross errors. This method assumes that the observed values change slowly over time and identifies the jump values in the observation process line. The observation time series is defined as $t_1$,…,$t_j$,…,$t_n$, with the corresponding observed values $y_1$,…,$y_j$,…,$y_n$. The jump characteristic for the $j$ measurement is defined as [33]

$$d_j=y_j-{\displaystyle \frac{y_{j+1}(t_j-t_{j-1})+y_{j-1}(t_{j+1}-t_j)}{t_{j+1}-t_{j-1}}}$$

(1)

If $y_j$ is a normal observed value, it only contains the linear deviation of the physical quantity’s time effect and random errors. When the number of observations n is sufficiently large, the average value of the jump characteristic $d_j$ will stabilize as follows:

$$\stackrel{-}{d}={\displaystyle \frac{1}{n-2}}{\displaystyle \sum _{j=2}^{n-1}|d_j}|$$

(2)

The jump standard deviation is

$$\sigma =\sqrt{{\displaystyle \sum _{j=2}^{n-1}{(|d_j|-\stackrel{-}{d})}^2/(n-3)}}$$

(3)

The necessary condition of outlier $y_j$ is

$$|d_j|-\stackrel{-}{d}\ge k\sigma$$

(4)

where k is an empirical coefficient, typically set to k = 2 or 3.

Taking the thermometer TJ37 at the foundation of dam section 11 and the strain gauge S3-1 at dam section 12 as the examples, the monitoring data during the construction period are selected, with no measurements during the winter shutdown period, as shown in Figure 6 and Figure 7. The marked points in the figures show significant jumps. Based on the comparisons with the monitoring instruments at the same location and for other dam sections, these jumps are preliminarily determined to be abnormal values. Additionally, using the jump feature detection method, the results indicate that both points are abnormal values. This method can accurately identify anomalies in the data, allowing for the removal of identified abnormal values to ensure the rationality of the monitoring data.

3. Simulation and Monitoring of Dam Temperature and Stress-Strain during Construction

In the simulation analysis of concrete dams, temperature serves as a fundamental load. The modeling process for dams in cold regions must prioritize considerations of target temperature, cooling water temperature, and the method of water cooling. This process involves adhering to a predetermined pouring schedule and basic temperature control measures, including surface protection via water cooling. Additionally, a sensitivity analysis of the temperature control parameters is conducted to study the distribution of a dam’s maximum temperature and maximum stress. By analyzing the characteristic points, the timing of the highest temperature and maximum stress occurrences can be determined, further enabling an analysis of the dam body’s stress variations. This approach allows for an examination of the temperature stress characteristics of the dam’s concrete in different areas and at different times under the influence of a cold region environment.

3.1. Basic Theories for Modeling Temperature Field and Stress Field

The differential equation of (5) for the temperature field of homogeneous and isotropic solids [34] is as follows:

$${\displaystyle \frac{{\partial }^{2}T}{{\partial x}^2}}+{\displaystyle \frac{{\partial }^{2}T}{{\partial y}^2}}+{\displaystyle \frac{{\partial }^{2}T}{{\partial z}^2}}+{\displaystyle \frac{1}{a}}\left({\displaystyle \frac{\partial \theta }{\partial \tau }}-{\displaystyle \frac{\partial T}{\partial \tau }}\right)=0$$

(5)

The boundary conditions are as follows:

$$T=\overline{T}$$

(6)

$$-\lambda {\displaystyle \frac{\partial T}{\partial n}}=q$$

(7)

$$-\lambda {\displaystyle \frac{\partial T}{\partial n}}=\beta \left(T-T_a\right)$$

(8)

where $\tau$ is the time (h); $\lambda$ is the thermal conductivity ($\mathrm{k}\mathrm{J}/\left(\mathrm{m}·\mathrm{h}·°\mathrm{C}\right)$); $\rho$ is the density ($\mathrm{k}\mathrm{g}/{\mathrm{m}}^3$); $c$ is the specific heat ($\mathrm{k}\mathrm{J}/\left(\mathrm{k}\mathrm{g}·°\mathrm{C}\right)$); $\alpha$ is the temperature transfer coefficient, $\mathsf{\alpha }=\mathsf{\lambda }/\mathrm{c}\mathrm{p}$(${\mathrm{m}}^2/\mathrm{h}$); $\theta$ is the adiabatic temperature rise (°C); $\overline{T}=\overline{T}\left(\tau \right)$ is the temperature (°C); $q=q\left(\tau \right)$ is the given heat flux ($\mathrm{k}\mathrm{J}/\left({\mathrm{m}}^2·\mathrm{h}\right)$); and $\beta$ is the surface heat release coefficient ($\mathrm{k}\mathrm{J}/\left({\mathrm{m}}^2·\mathrm{h}·°\mathrm{C}\right)$). Under the condition of natural convection, $T_a$ is the air temperature; under forced convection, $T_a$ is the adiabatic wall temperature of the boundary layer.

In the simulation analysis of a concrete dam, the temperature is the basic load. The solution of the transient temperature field is the temperature field function $T=T_0\left(x,y,z\right)$ that satisfies the transient heat conduction equation and boundary conditions under the initial condition of $T\left(x,y,z,\tau \right)$.

If $\overline{T}\left(\tau \right)$, $q\left(\tau \right)$, $T_a$ and $\theta$ on the boundary do not change with time. Then, after a certain time of heat exchange, the temperature field in the object will not change with time, that is, ${\displaystyle \frac{\partial T}{\partial \tau }}=0$; the transient heat conduction equation degenerates into the steady-state heat conduction equation, and $\mathrm{T}$ is only related to the coordinates.

According to the principle of minimum potential energy, the heat conduction differential Equation (5) can be transformed into the extremum problem of the functional (9), with the temperature $T\left(x,y,z,\tau \right)$ at $\tau =0$ given the initial temperature $T_0\left(x,y,z\right)$, and $\overline{T}\left(\tau \right)$ given the boundary condition at boundary $C_1$:

$$\begin{gathered} I(T)={\displaystyle {\iiint }_R\left\{\frac{1}{2}\left[{(\frac{\partial T}{\partial x})}^2+{(\frac{\partial T}{\partial y})}^2+{(\frac{\partial T}{\partial z})}^2\right]+\frac{1}{a}(\frac{\partial T}{\partial \tau }-\frac{\partial \theta }{\partial \tau })T\right\}dx}dydz+{\displaystyle {\iint }_{C_2}\overline{q}Tds+} \\ {\displaystyle {\iint }_{C_3}(\frac{\overline{\beta }}{2}{T}^2-\overline{\beta }{T}_{a}T)ds} \end{gathered}$$

(9)

where $\overline{\beta }=\beta /\lambda$, $\overline{q}=q/\lambda$, $\theta ={\theta }_0\left(1-e^{-m_1{\tau }^{m2}}\right)$, and ${\theta }_0$ is the final adiabatic temperature rise of the concrete.

After the concrete temperature field $\mathrm{T}$ is solved, the temperature stress of each part needs to be further calculated. Temperature deformation only produces linear strain and does not produce shear strain. This linear strain can be regarded as the initial strain of the object. When calculating the temperature stress, the deformation ${\epsilon }_0$ caused by the temperature is calculated first, and then the equivalent node temperature load $P_{{\epsilon }_0}$ caused by the corresponding initial strain is obtained. Then, the node displacement caused by the temperature change is obtained according to the usual solution stress method, and then the temperature stress $\sigma$ is obtained. The equivalent node temperature load $P_{{\epsilon }_0}^e$ of element e is as follows:

$$P_{{\epsilon }_0}^e={\displaystyle {\iiint }_{\Delta R}{B}^{T}D{\epsilon }_{0}dR}$$

(10)

where $B$ is the transformation matrix of the strain and displacement, $D$ is the elastic matrix, and R is any direction in unit volume. The equivalent node load $P_{{\epsilon }_0}$ caused by temperature deformation can be added together with the other load terms to obtain the total stress, including the temperature stress.

The stress–strain relationship of the calculated stress includes the initial strain term

$$\mathsf{\sigma }=\mathrm{D}\left({\mathsf{\epsilon }-\mathsf{\epsilon }}_0\right)$$

(11)

Concrete is an elastic creep body, and the creep effect $C\left(t,\tau \right)$ of the concrete needs to be considered in the simulation calculation process. Using the incremental method, the time $\tau$ is divided into a series of time periods: $∆{\tau }_1$, $∆{\tau }_2$,…, $∆{\tau }_n$. The relationship between the stress increment and strain increment is as follows:

$$\left\{∆{\mathsf{\sigma }}_{\mathrm{n}}\right\}=\left[\overline{D_n}\right]\left(\left\{∆{\epsilon }_n\right\}-\left\{{\eta }_n\right\}-\left\{∆{\epsilon }_n^T\right\}-\left\{∆{\epsilon }_n^0\right\}-\left\{∆{\epsilon }_n^s\right\}\right)$$

(12)

where $\left\{∆{\epsilon }_n\right\}$ is the strain increment generated in the $∆{\tau }_n$ period; $\left\{{\eta }_n\right\}$ is the displacement increment; $\left\{∆{\epsilon }_n^T\right\}$ is the temperature strain increment; $\left\{∆{\epsilon }_n^0\right\}$ is the autogenous volume deformation increment; and $\left\{∆{\epsilon }_n^s\right\}$ is the dry shrinkage strain increment.

The following formula are used:

$$\left[\overline{D_n}\right]=\overline{E_n}{\left[Q\right]}^{-1}$$

(13)

$$\overline{E_n}={\displaystyle \frac{E\left(\overline{{\tau }_n}\right)}{1+E\left(\overline{{\tau }_n}\right)C\left({\tau }_n,\overline{{\tau }_n}\right)}}$$

(14)

The stress increment $\left\{∆{\sigma }_n\right\}$ is obtained from (12), and the stress of each element ${\tau }_n$ is obtained after accumulation.

$$\left\{{\mathsf{\sigma }}_{\mathrm{n}}\right\}=\sum \left\{∆{\mathsf{\sigma }}_{\mathrm{n}}\right\}$$

(15)

3.2. Cooling Water Pipe Simulation

In cold climate conditions, it is essential to fully consider the impact of the dam body’s temperature control measures on the simulated conditions, especially the thermal conductivity effect of the cooling pipe network. This paper takes into account various factors, including the different target temperatures, different cooling water temperatures, and different water flow cooling methods, according to the proposed pouring schedule and basic temperature control measures, to carry out a multi-condition simulation analysis.

Considering the cooling problem of a single water pipe, the diameter of the concrete cylinder is D, the length is L, the initial temperature of the concrete is T₀, and the cooling water temperature at the inlet of the water pipe is T_w. Assuming that the adiabatic temperature rise of the concrete is $\theta \left(\tau \right)$, the cooling water pipe is regarded as a negative heat source, and the effect of the cooling water pipe is considered in the average sense. The equivalent heat conduction equation of the concrete can be obtained as follows:

$${\displaystyle \frac{\partial \mathrm{T}}{\partial t}}=a{\nabla }^{2}T+(T_0-T_w){\displaystyle \frac{\partial \mathsf{\Phi }}{\partial t}}+{\theta }_0{\displaystyle \frac{\partial \mathsf{\psi }}{\partial t}}$$

(16)

where $\tau$ is the time, $\mathrm{a}$ is the temperature transfer coefficient, ${\nabla }^{2}T$ is the Laplace operator, $\Phi \left(t\right)$ and $\psi \left(t\right)$ are the cooling effect functions of the water pipe, and T is the average temperature of the concrete.

Currently, the cooling water pipes used in hydroelectric dam projects are primarily of two types: metal pipes and plastic pipes. Extensive theoretical analyses, simulation calculations, and engineering application results have shown that, under identical conditions, metal water pipes with an outer diameter of 28.4 mm and an inner diameter of 25.4 mm have essentially the same cooling effect as plastic water pipes with an outer diameter of 32 mm and an inner diameter of 28 mm. The difference in the cooling effectiveness between metal and plastic water pipes is approximately 0.4 to 0.5 $°\mathrm{C}$.

3.3. Numerical Model and Model Parameters

This paper primarily simulates the non-overflow dam section. The computational model and mesh for the non-overflow dam section are shown in Figure 8a. The model consists of 541,336 elements and 565,073 nodes. In the stress calculations, the rock foundation bottom surface is constrained in three directions; the top surface is free; and the left, right, upstream, and downstream surfaces are constrained in the normal axial direction. The four sides and the top surface of the dam body are free. The partitioning of the dam concrete materials is shown in Figure 8b. The dam concrete construction is carried out by block pouring. For this project, section 12# of the dam is divided into 17 blocks for pouring, with summer pouring temperatures controlled at 15 °C, according to the temperature control measures. The dam pouring schedule is shown in

Table 1. Dam pouring progress.

Warehouse Number	Pouring Time	Pouring Height	Pouring Temperature
1	10 April 2022	1.0 m	Natural Pouring Temperature
2	27 April 2022	1.5 m	Natural Pouring Temperature
3	29 May 2022	1.5 m	Natural Pouring Temperature
4	4 June 2022	1.5 m	Control Pouring Temperature
5	11 June 2022	1.5 m	Control Pouring Temperature
6	23 June 2022	1.5 m	Control Pouring Temperature
7	2 July 2022	1.5 m	Control Pouring Temperature
8	7 July 2022	1.5 m	Control Pouring Temperature
9	13 July 2022	1.5 m	Control Pouring Temperature
10	24 March 2023	3.4 m	Natural Pouring Temperature
11	4 April 2023	3.0 m	Natural Pouring Temperature
12	13 May 2023	3.0 m	Natural Pouring Temperature
13	25 May 2023	3.0 m	Natural Pouring Temperature
14	4 June 2023	3.0 m	Control Pouring Temperature
15	14 June 2023	3.0 m	Control Pouring Temperature
16	24 June 2023	1.8 m	Control Pouring Temperature
17	30 June 2023	1.2 m	Control Pouring Temperature

.

The simulation employs the birth-and-death element method. Initially, all the dam elements are “killed”. Then, the elements are gradually “activated” according to the actual construction schedule. This process includes simulating the heat exchange, cooling measures, and construction breaks to analyze the transient temperature field of the dam. Subsequently, the element types are changed, material parameters are assigned, and the temperature field results are imported into the structural analysis as volumetric loads. Boundary conditions and other loads are then applied to perform a transient stress field analysis.

The concrete mix ratio is one of the key aspects of production and construction, playing a significant role in ensuring project quality. Twenty-four different mix proportions were formulated and tested, each with varying water-cement ratios and fly ash contents. A comprehensive analysis of the concrete’s strength, frost resistance, and permeability test results were conducted. Based on these analyses, appropriate concrete mix proportions were selected for the different material zones of the dam. The concrete mix proportions for the dam are shown in

Table 2. Dam concrete mix ratio.

Design Strength	Gradation	Water-Binder Ratio	Fly Ash (%)	Percentage of Sand (%)	Water Reducer (%)	Air- Entraining Agent (1/10 k)	Concrete Material Consumption (kg/m3)
							Water	Cement	Fly Ash	Sand	Stone
C25	two	0.35	15	38	0.50	3.2	128	311	55	674	1147
C30	two	0.35	25	38	0.45	3.2	126	270	90	670	1140
C15	three	0.52	30	36	0.50	2.8	102	137.3	58.8	723	1345
C25	three	0.40	30	34	0.45	3.6	102	179	77	657	1336

.

The selection of the model parameters is critical for accurately reflecting real-world conditions in simulations. Based on the characteristics of the cold climate conditions at the dam site and the experimental research results of this project, the boundary conditions for cold region concrete dams, concrete materials, bedrock, length of pouring blocks, and dam body positions were determined. The concrete performance parameter information was derived from the indoor test data, as shown in the following

Table 3. Thermodynamic parameters of concrete.

Temperature Transfer Coefficient (m2/h)	Thermal Conductivity (kJ/m·h·°C)	Specific Heat (kJ/kg·°C)	Thermal Expansion Coefficient (10−6/°C)	Density (g/cm3)	Poisson Ratio
0.0024	5.46	0.96	9.0	2.39	0.20

.

The tensile strength and compressive strength of the dam concrete are shown in

Table 4. Tensile strength and compressive strength of dam concrete.

Concrete Position	Tensile Strength (MPa)				Compressive Strength (MPa)
Age	7 d	28 d	90 d	180 d	7 d	28 d	90 d
Dam Interior	0.82	1.39	1.86	2.23	14.3	21.3	29.2

.

The fitting of the time-dependent elastic modulus curves for the main body of the dam and the various concrete regions utilizes Equation (17). The fitted elastic moduli for the different types of concrete are presented in

Table 5. Modulus of elasticity of concrete.

Concrete Partition	${\mathbf{E}}_0$ (GPa)	a	b
Dam Body	42.7	0.476	0.293
Cushion	43.7	0.476	0.293
Upstream Surface	43.7	0.476	0.293
Downstream Surface	43.7	0.476	0.293

. Here, $E_0$ is the final elastic modulus of the concrete, and a and b are determined by fitting the experimental data.

$$E_c\left(t\right)=E_0\left(1-e^{-{at}^b}\right)$$

(17)

Temporary insulation is applied to the exposed concrete surfaces during the construction period, using 2 cm thick polyurethane foam blankets with an equivalent heat release coefficient of 98.58 kJ/(m²·d·°C). For the permanently exposed surfaces, such as the upstream and downstream faces, temporary insulation is used before October, and permanent insulation is applied after October. After insulation, the equivalent heat exchange coefficient for the upstream concrete surface is 23.9 kJ/(m²·d·°C), and for the downstream concrete surface it is 19.96 kJ/(m²·d·°C). The heat exchange coefficient for the temporary heat dissipation surfaces of the concrete is 200 kJ/(m²·d·°C).

The water cooling for the dam body in this project adopts a two-stage cooling method. For each layer of concrete poured between May and August, water pipes are arranged at intervals of 1.5 m × 1.5 m (vertical spacing × horizontal spacing) for water cooling, with the water temperature maintained at 15 °C. The layout is shown in Figure 9. The cooling water pipes are high-strength polyethylene (HDPE) pipes with an inner diameter of 32.6 mm and an outer diameter of 40 mm. The density of the HDPE pipes is 940 kg/m³, the thermal conductivity is 0.4 W/m·K, and the specific heat capacity is 2301 J/kg·K.

The second stage of dam cooling is conducted in early November, with the cooling water temperature set at 10 °C and the target temperature at 22 °C. The first stage cooling flow rate is 1.8 m³/h, while the second stage cooling flow rate is 1.2 m³/h.

3.4. Dam Temperature Analysis

Based on the results of the thermodynamic parameter tests of the concrete, a simulation of the dam body temperature during the construction and operation periods was carried out. The contour maps of the highest temperature in the downstream profile of the concrete are shown in Figure 10, Figure 11, Figure 12 and Figure 13 below. From the temperature distribution contour plot for the construction period, it can be seen that during the construction phase of the concrete gravity dam, as the concrete is continuously poured layer by layer, the highest temperature region in the dam body appears at the center of the poured layers. This is due to the large volume of poured concrete, which has less heat exchange compared to the surface areas of the dam, causing heat to accumulate inside the dam body and resulting in a significant temperature rise. When the dam body is poured to an elevation of 120.02 m, the pouring height is 4 m. At this point, the concrete is located in the strongly constrained foundation area, and the highest temperature of the dam body is 33.42 °C, appearing inside the concrete at an elevation of 119.5 m. When the dam body is poured to an elevation of 126.02 m, with a pouring height of 10 m, the concrete is in the weakly constrained foundation area, and the highest temperature of the dam body is 36.01 °C, appearing inside the concrete at an elevation of 125.5 m.

After implementing the water cooling measures for the concrete poured during the high-temperature season from May to August, the calculated maximum internal temperature of the concrete approaches 35 °C. This indicates that the dam body temperature is closely related to the pouring season. For the internal concrete, the characteristic point temperatures near the surface, influenced by the atmospheric temperature after the end of water cooling, show annual fluctuations. The temperature fluctuation amplitude gradually decreases for the concrete near the interior of the dam body. For the upstream surface concrete, the hydration heat gradually dissipates after pouring, causing the concrete temperature to rise to a peak and then gradually decline. After applying permanent insulation measures in October, the concrete surface temperature rises slightly, eventually fluctuating with the atmospheric temperature. The cooling water pipes serve as a negative heat source. After controlling and reducing the temperature within the dam body, the water temperature rises, as shown in Figure 14.

At an elevation of 117.02 m inside the dam body of section 12#, four thermometers (TT75–TT78) were installed. TT75 is located near the upstream heel of the dam, and TT78 is near the downstream toe of the dam; both were installed using a post-embedding method. TT76 and TT77 are located inside the dam body and were installed using a pre-embedding method. Due to the length of the monitoring data sequence, this paper only analyzes the construction period temperature data from the four thermometers TT75–TT78. The monitoring data and numerical simulation results are shown in Figure 15, Figure 16, Figure 17 and Figure 18 below.

Through a comparative analysis, it can be seen that the simulation results reasonably reflect the temperature of the dam during the construction period. After the upper layer of concrete is poured, the heat generated from the hydration reaction diffuses into the lower layer, causing a certain temperature rise in the lower concrete. The temperature of the dam concrete primarily goes through two stages. In the first stage, after each section of the dam is completed, the temperature gradually increases, reaching its peak in about 7 days. In the second stage, under the influence of the external low-temperature environment and temperature control measures, the dam temperature gradually decreases and exhibits periodic fluctuations in response to changes in the external temperature, with a certain lag compared to the ambient temperature. Additionally, the temperature at the internal measurement points within the dam is relatively higher, while the temperature near the upstream and downstream dam surfaces is lower. Due to the duration of direct sunlight exposure, the upstream dam face, being on the shaded side, has the lowest measured temperature.

The thermodynamic material parameters used in this simulation are based on laboratory test results, which differ somewhat from the actual thermal conductivity of the on-site materials. As a result, the simulated initial temperature of the concrete pouring is higher than that of the monitored results. In the initial phase of dam pouring, the temperature control measures are effective, leading to an average dam body temperature lower than that of the simulated temperature. As the influence of the ambient temperature increases, the dam body temperature gradually aligns with the simulated temperature. Additionally, due to the presence of construction water accumulation in the upstream face padding area during the construction period, the temperature at the TT75 thermometer location is influenced by the water temperature. This results in lower measurements at the beginning of the 2023 construction period, causing some discrepancies between the monitored and simulated data.

3.5. Dam Stress-Strain Analysis

A stress simulation of the dam body during the construction and operation periods was conducted, and the contour plots of the maximum stress and maximum principal stress for the non-overflow dam section are shown in Figure 19 and Figure 20 below. In the simulation results, the compressive stress is positive and the tensile stress is negative, with the unit being Pa. Based on the range of the stress influence, due to the dam body being constrained by the foundation rock, the foundation constraint area is within a height of 0.4 L of the long side dimension of the pour block, where the strong constraint area of the foundation is 0–0.2 L, and the weak constraint area is 0.2–0.4 L.

The simulation results indicate that the maximum tensile stress inside the main body of concrete appeared at the contact surface of the poured layers. The maximum tensile stress during the 2022 construction period was 3.31 MPa, and during the 2023 construction period it was 4.99 MPa. Additionally, a significant tensile stress was generated in the padding area of the strong constraint zone. The average tensile stress in the padding area of the strong constraint zone was approximately 2.03 MPa during the 2022 construction period, and about 1.58 MPa during the 2023 construction period. The compressive stress at the upstream and downstream faces of the dam body is greater than the internal compressive stress of the dam body, with stress concentration occurring at the heel of the dam, resulting in a high compressive stress.

The initial curing period (24 h) measurements of the strain gauge groups, which no longer fluctuate and whose resistance ratios and resistances correspond to the final setting moment, are taken as the baseline values. According to the theory of elastic deformation, at any point within the dam body, the sum of strains ε1, ε2, and ε3 in three mutually perpendicular directions of a five-direction strain gauge is constant. The strain gauge groups calculate stress in either a plane or in space, where the sum of directions 1 and 3 equals the sum of directions 2 and 4, that is, Sn−1 + Sn−3 = Sn−2 + Sn−4. The observation period for the strain gauges was from 21 June 2022 to 31 October 2023, with the monitoring results shown in Figure 21, Figure 22, Figure 23, Figure 24 and Figure 25.

The actual stress of the concrete is solved by the deformation method in [29]. Assuming that the age to be solved for falls within the interval $\left(t_i,t_{i+1}\right)$ of the test age, that is, $E\left(t_i\right)$ and $E\left(t_{i+1}\right)$ are known, the semi-logarithmic linear interpolation can be obtained

$$E\left(\tau \right)=E\left(t_i\right)+\left(ln\tau -{lnt}_i\right){\displaystyle \frac{E\left(t_{i+1}\right)-E\left(t_i\right)}{{lnt}_{i+1}-{lnt}_i}}$$

(18)

The creep degree is related to the history of stress action. It is assumed that the loading age $\mathsf{\tau }$ falls in the test age $\left({\tau }_i,{\tau }_{i+1}\right)$ interval, and the age $\mathrm{t}$ falls in the test age $\left(t_j,t_{j+1}\right)$ interval. The creep of the concrete at the loading age of 3 days, 14 days, 28 days, and 90 days is obtained through laboratory tests. The creep of the different periods can be obtained by a continuous semi-logarithmic linear interpolation:

$$X=c\left(t_i,{\tau }_i\right)+\left(lnt-{lnt}_j\right){\displaystyle \frac{c\left(t_{j+1},{\tau }_i\right)-c\left(t_j,{\tau }_i\right)}{{lnt}_{j+1}-{lnt}_j}}$$

(19)

$$Y=c\left(t_i,{\tau }_{i+1}\right)+\left(lnt-{lnt}_j\right){\displaystyle \frac{c\left(t_{j+1},{\tau }_{i+1}\right)-c\left(t_j,{\tau }_{i+1}\right)}{{lnt}_{j+1}-{lnt}_j}}$$

(20)

$$c\left(t,\tau \right)=X+\left(ln\tau -{ln\tau }_i\right){\displaystyle \frac{Y-X}{{ln\tau }_{i+1}-{ln\tau }_i}}$$

(21)

The uniaxial strain calculation formula is as follows:

$${\epsilon }_x^{\prime }={\displaystyle \frac{{\epsilon }_x}{1+\mu }}+{\displaystyle \frac{\mu }{\left(1+\mu \right)\left(1-2\mu \right)}}\left({\epsilon }_x+{\epsilon }_y+{\epsilon }_z\right)$$

(22)

$${\epsilon }_y^{\prime }={\displaystyle \frac{{\epsilon }_y}{1+\mu }}+{\displaystyle \frac{\mu }{\left(1+\mu \right)\left(1-2\mu \right)}}\left({\epsilon }_x+{\epsilon }_y+{\epsilon }_z\right)$$

(23)

$${\epsilon }_z^{\prime }={\displaystyle \frac{{\epsilon }_z}{1+\mu }}+{\displaystyle \frac{\mu }{\left(1+\mu \right)\left(1-2\mu \right)}}\left({\epsilon }_x+{\epsilon }_y+{\epsilon }_z\right)$$

(24)

Based on the calculated uniaxial strain, combined with the concrete creep test results and the elastic modulus test results, the concrete stress is calculated by the deformation method. The time is divided into n periods, and the starting and ending time (age) of each period are ${\tau }_0$, ${\tau }_1$,…, ${\tau }_{i-1}$, ${\tau }_i$,…, ${\tau }_{n-1}$, ${\tau }_n$, respectively. The midpoint age of each period is ${\overline{\tau }}_i=\left({\tau }_i+{\tau }_{i-1}\right)/2$. The uniaxial strain corresponding to each moment is ${\epsilon }_0^{\prime }$, ${\epsilon }_1^{\prime }$,…, ${\epsilon }_i^{\prime }$,…, ${\epsilon }_n^{\prime }$. The uniaxial strain increment for each period is ${{\Delta \epsilon }_i^{\prime }=\epsilon }_i^{\prime }-{\epsilon }_{i-1}^{\prime }$. The actual stress of the concrete at time ${\tau }_n$ is calculated as follows:

$$\sigma \left({\overline{\tau }}_n\right)=\sum _{i=1}^n\Delta \sigma \left({\overline{\tau }}_i\right)$$

(25)

$\sigma \left({\overline{\tau }}_i\right)$ is the stress increment at time ${\overline{\tau }}_i$, and is calculated as follows:

$$\Delta \sigma \left({\overline{\tau }}_i\right)=E^{\prime }\left({\overline{\tau }}_i,{\tau }_{i-1}\right)\overline{{\epsilon }_i^{\prime }}(\mathrm{i}=1)$$

(26)

$$\Delta \sigma \left({\overline{\tau }}_i\right)=E^{\prime }\left({\overline{\tau }}_i,{\tau }_{i-1}\right)\left\{\overline{{\epsilon }_i^{\prime }}-\sum _{j=1}^{i-1}\Delta \sigma \left({\overline{\tau }}_j\right)\times \left[{\displaystyle \frac{1}{E\left({\tau }_{j-1}\right)}}+c\left({\overline{\tau }}_i,{\tau }_{j-1}\right)\right]\right\}(\mathrm{i}>1)$$

(27)

In the formula, $E^{\prime }\left({\overline{\tau }}_i,{\tau }_{i-1}\right)$ is the reciprocal of the total deformation $\left[{\displaystyle \frac{1}{E\left({\tau }_{j-1}\right)}}+c\left({\overline{\tau }}_i,{\tau }_{j-1}\right)\right]$ when the unit stress of the ${\tau }_{i-1}$ age loading continues to ${\overline{\tau }}_i$, which is called the continuous elastic modulus at ${\overline{\tau }}_i$ time. $E\left({\tau }_{j-1}\right)$ is the instantaneous elastic modulus of concrete at ${\tau }_{j-1}$ time. $c\left({\overline{\tau }}_i,{\tau }_{j-1}\right)$ takes ${\tau }_{j-1}$ as the creep degree when the loading age continues to ${\overline{\tau }}_i$. The stress calculation results are shown in

Table 6. Statistical table of characteristic values of extreme stress (maximum, minimum) of five-direction strain gauges.

Time	Stress Directions (MPa)	Instrument Number
		S1	S2	S3	S4	S5
Stress distribution map of the dam during the 2022 construction period.	Left and Right Banks	0.87/−2.91	1.80/−1.55	2.08/0.20	*	*
	Vertical Direction	0.44/−3.14	0.60/−0.82	1.73/−0.23	*	*
	Upstream and Downstream	0.98/−2.93	0.87/−1.10	6.20/−0.18	*	*
Stress distribution map of the dam during the 2023 construction period.	Left and Right Banks	4.03/−1.12	2.94/−0.78	3.05/−2.26	0.10/−0.44	0.05/−2.59
	Vertical Direction	3.86/−1.51	2.58/−0.73	1.99/−8.11	0.10/−0.10	0.06/−0.01
	Upstream and Downstream	3.68/−1.87	2.55/−0.58	4.97/−2.77	0.36/−0.66	0.01/−0.01

below. The S4 and S5 five-way strain gauges were installed and embedded in 2023, so there is no data for the 2022 construction period, indicated by *.

From the stress extremum statistics table, it can be seen that the maximum tensile stress in the left and right bank directions of the dam is 4.03 MPa, which occurred at the beginning of the 2023 construction period. The maximum tensile stress in the upstream and downstream directions is 6.2 MPa, which occurred during the 2022 construction period. The maximum tensile stress in the vertical direction is 3.86 MPa, which also occurred at the beginning of the 2023 construction period. The environmental temperature significantly affected the stress on the dam surface. Due to the influence of the low-temperature environment during the winter shutdown period of 2022, the dam body experienced significant tensile stress, with the tensile stress on the upstream and downstream faces being greater than that inside the dam body. During the construction period, as the dam pouring progressed, the compressive stress at the monitoring points gradually increased with the pouring of the upper dam body. The areas with higher stress values were mainly concentrated at the heel and toe of the dam. After the dam pouring was completed, the interior of the dam body was primarily in a state of compression, with smaller tensile stresses present at the upstream and downstream faces. The compressive stress in the dam did not exceed the compressive strength of the concrete, while the tensile stress in certain areas exceeded the tensile strength. The analysis suggests that during the strain-to-stress conversion, the creep data used were from the experiments, which differed to some extent from the actual creep behavior of the on-site materials. Additionally, due to the concrete’s toughness and ductility during the initial stress increase, and the fact that the stress in localized areas did not reach the material’s tensile limit, no cracks appeared in the dam.

During the dam pouring process, there was a significant temperature difference between the upper and lower layers of the poured concrete. The longer the pouring interval, the more fully the concrete dissipated heat, resulting in a lower temperature rise and a lower maximum principal stress. Therefore, reasonably controlling the dam’s concrete pouring schedule could effectively mitigate the tensile stress between the upper and lower pouring layers of the dam body.

Additionally, due to the high maximum temperature and low winter temperatures, the temperature drop in the padding area concrete could reach up to 20 °C, leading to increased tensile stress. It is necessary to adopt stricter temperature control measures to reduce the temperature rise, as well as material optimization measures, such as decreasing the adiabatic temperature rise and the linear expansion coefficient of the concrete, to alleviate surface temperature stress in this area. Given the dam’s location in a cold region, winter insulation measures were applied to the dam body during the winter shutdown period. However, upon resuming construction, significant tensile stress still occurred at the overwintering surface. The overwintering surface experienced noticeable tensile stress due to the combined effects of the temperature difference between the upper and lower layers, and the internal and external temperature differences, consistent with the monitoring results.

4. Discussion

4.1. Comparison and Analysis of Temperature Control Measures for the Dam

Concrete gravity dams rely on their own weight for stability, resulting in a large volume and high cement usage, which leads to significant amounts of final hydration heat. This can cause considerable temperature and shrinkage stresses during construction, with the temperature stress constituting a significant portion of the total stress. Thus, it is crucial to strictly control the temperature conditions during construction. By comparing the temperature control measures of the Guanmenzuizi Reservoir Dam with those of typical Chinese concrete gravity dams [35,36], the rationality of the temperature control measures implemented in the Guanmenzuizi Reservoir Dam is analyzed, and the conditions for optimizing these measures are explored. The typical temperature control measures for dams are presented in

Table 7. Temperature control measures for typical concrete gravity dam.

Project Name	Position	Raw Material	Water Cooling	Dam Face Protection	Placing Temperature Control
Guanmenzuizi Dam	cold region	The strength grade is 42.5 ordinary Portland cement.	Second-stage water cooling	temporary and permanent insulation	Summer pouring temperature ≤ 15 °C.
Three Gorges Dam	Non-cold region	The strength grade of micro-expansion performance is 42.5 medium-heat cement.	Second-stage water cooling	temporary and permanent insulation	In addition to the winter construction period, the foundation constraint region is 12–14 °C; 16–18 °C in other regions.
Zangmu Dam	cold region	The strength grade is 42.5 medium-heat Portland cement.	Third-stage water cooling	temporary and permanent insulation	Outlet temperature ≥ 10 °C from November to early March; from May to September, the strong constraint region was ≤10 °C, and the other regions were ≤12 °C.

below.

Based on the comprehensive comparison of factors, such as the ambient temperature at the dam site, the dam construction materials, and the dam temperature control measures, the feasibility of optimizing the temperature control measures is proposed. For the Three Gorges Dam, it was suggested that the concrete in the horizontal tensile zone of the dam body was more extensive because the foundation constraint area concrete was poured in summer. Therefore, by optimizing the construction schedule, the concrete in the foundation constraint area of a dam can avoid being poured in summer. The Three Gorges Dam and the Zangmu Hydropower Station dam selected specific cement for the dams’ construction. By adopting material optimization measures, such as reducing the adiabatic temperature rise and the coefficient of the linear expansion of concrete, the temperature stress on the upstream and downstream surfaces of a dam body can be alleviated. Ordinary Portland cement is used in this project.

Scholars, such as Yang et al., have studied the layout of the cooling water pipes for the Xiangjiaba concrete gravity dam and determined the impact of the spacing of the cooling water pipes on the dam’s concrete [37]. The density of the cooling water pipes’ layout affects both the compressive and tensile stresses of the dam body; the higher the density, the smaller the internal compressive stress of the dam body. However, when the density is too high, it can cause significant tensile stress on the dam surface. In this project, the density of the cooling water pipes’ layout is relatively high. Therefore, optimizing the density of the cooling water pipes’ layout can be considered as a way to reduce the tensile stress on the upstream and downstream surfaces of the dam body.

Through a comparative analysis with other typical concrete gravity dams, this project adopted strict temperature control measures. The layout of the various measures was relatively reasonable, and specific treatments were implemented for the low-temperature environments of cold regions. The temperature of the dam concrete was effectively controlled, providing a reference for the selection of temperature control measures for similar concrete gravity dams in cold regions.

4.2. Analysis of the Effects of the Spacing of Cooling Water Pipes

Through a comparison with other typical cooling water pipe layouts for concrete gravity dams, this study explores the optimization scheme for the cooling water pipe layout of the Guanmenzuizi Dam. The horizontal and vertical spacing of the cooling water pipes are adjusted separately. Scenario one adjusts the vertical spacing to 2 m, and scenario two adjusts the horizontal spacing to 2 m. The temperature field and stress field distribution of the dam body are simulated for this analysis.

(1)

Temperature Field

Comparing the two scenarios with the actual cooling water pipe layout of the dam, the effects of different cooling water pipe arrangements on the dam body temperature are analyzed. The temperature fields are shown in Figure 26, Figure 27, Figure 28 and Figure 29. From the comparison of the temperature fields of the different cooling water pipe layout scenarios, it can be seen that the highest temperature in the dam body for scenario one is 41.21 °C, while for scenario two it is 40.86 °C. The impact of increasing the vertical spacing on the temperature rise in the dam body is greater than that of increasing the horizontal spacing. The dam body temperature is lowest in scenario two, indicating that the temperature control effect of scenario two is better than the actual scenario and scenario one. Therefore, adjusting the horizontal spacing to 2 m provides the best cooling effect for the dam body.

(2)

Stress Field

By comparing the stress fields of the two scenarios with the actual cooling water pipe layout of the dam, the effects of the different cooling water pipe arrangements on the dam body stress are analyzed. The stress fields are shown in Figure 30, Figure 31, Figure 32, Figure 33, Figure 34, Figure 35, Figure 36 and Figure 37. From the comparison of the stress fields of different schemes, the following can be seen:

For scheme one, the maximum tensile stress during the 2022 construction period is 3.32 MPa, and the maximum tensile stress in 2023 is 4.99 MPa. For scheme two, the maximum tensile stress during the 2022 construction period is 3.33 MPa, and the maximum tensile stress in 2023 is 5.03 MPa.

During the 2022 construction period, the tensile stress of the dam body in the actual scenario is slightly lower than that of scenario one. After the dam body is completed, the tensile stress of the dam body in scenario one is slightly lower than that of the actual scenario. Both scenario one and the actual scenario are more effective at controlling stress compared to scenario two, indicating that the actual scenario can effectively reduce the tensile stress of the dam body.

By comparing the temperature and stress field results of the two scenarios with the actual scenario, it is found that scenario two can more effectively control the dam body temperature, but results in the highest stress. Scenario one’s effects are similar to those of the actual scenario. The actual cooling water pipe layout scenario of the dam is reasonable and effectively controls the generated tensile stresses in the dam.

5. Conclusions

This study investigates the feasibility of temperature control measures during the construction of the Guanmenzuizi Reservoir concrete dam. It conducts finite element simulations of the dam body’s temperature and stress under conditions that include temperature control measures, and validates these simulations with the field monitoring results of the temperature and stress-strain. The main conclusions are summarized as follows:

(1)

There are temperature variation patterns in the dam concrete during the construction period. After the upper layer of concrete is poured, the heat from the hydration reaction diffuses into the lower layer, causing a certain temperature rise in the lower concrete. The temperature of the dam concrete generally goes through two stages. In the first stage, after the completion of each section, the temperature gradually increases and reaches its peak in about 7 days. In the second stage, influenced by the external low-temperature environment and the temperature control measures, the dam temperature gradually decreases and exhibits periodic fluctuations in response to the external temperature changes, with a certain lag compared to the ambient temperature. Additionally, the temperature at the internal measurement points within the dam is relatively higher, while the temperature at the points near the upstream and downstream dam surfaces is lower. Due to the duration of direct sunlight exposure, the upstream dam face, being on the shaded side, has the lowest measured temperature.

(2)

During the construction period of the dam, the compressive stress gradually increased as the upper layers were poured, with the high-stress areas mainly concentrated at the heel and toe of the dam. In cold regions, the low winter temperatures cause significant tensile stress on the upstream and downstream surfaces of concrete dams. Therefore, it is necessary to implement effective temperature control measures to reduce the rise in the concrete’s temperature, or to optimize the dam materials by reducing the adiabatic temperature rise and the coefficient of the linear expansion of the concrete, thereby lowering the temperature-induced stress on the dam’s surface. In this project, measures, such as controlling the pouring temperature during summer, surface spraying, surface water flow, and phase-one water cooling, effectively reduced the early temperature rise of the dam. In response to the low winter temperatures of this cold region, secondary cooling and dam surface insulation measures were taken to reduce the temperature difference between the interior and exterior of the dam. Numerical simulations and on-site monitoring results show that the temperature control measures adopted in this project are reasonable, providing a valuable reference for temperature control during the construction of concrete gravity dams in cold regions.

(3)

Based on a comparison with the temperature control measures used during the construction period of other typical concrete gravity dams in China, the rationality of the temperature control measures used for the Guanmenzuizi Dam is analyzed. The results indicate that this project implemented strict temperature control measures, which were reasonably arranged and specifically addressed the low-temperature environment of cold regions. It is suggested that the thermal rise and linear expansion coefficient of concrete can be reduced by optimizing the dam materials. Alternatively, the tensile stress on the dam body can be reduced by optimizing the construction organization design for concrete pouring, avoiding pouring the dam foundation constrained area in the summer. This study also explores the impact of the cooling water pipe density on the dam body. Comparing two scenarios with the actual cooling water pipe scenario, the actual layout of the cooling water pipes is found to be reasonable and to effectively control the tensile stress of the dam body.