Finite Element Method Simulation and Experimental Investigation on the Temperature Control System with Groundwater Circulation in Bridge Deck Pavement

Abstract

The application of green energy resources is gaining increasing attention in the field of engineering. In cold areas, the groundwater circulation temperature control system (GCTCS) can serve as an auxiliary structure to the bridge deck on highways, effectively preventing the pavement surface from freezing. In this study, a finite element simulation is conducted to establish a bridge structure model of the GCTCS, incorporating both steady-state and transient conditions to investigate its anti-icing performance. Additionally, the influences of various factors, such as wind speed, asphalt concrete layer thickness, groundwater temperature, pipe water flow rate, and pipe spacing, on the temperature of the water film on the pavement surface are investigated and validated through laboratory testing. The results demonstrate that wind speed has a significant influence, with the convective heat loss reaching 90% when the wind speed reaches 10 m/s. Groundwater temperature is the second most influential factor, showing a linear relationship with the water film temperature. Excessive pipe spacing can lead to an uneven temperature distribution on the pavement surface. The thickness of the asphalt concrete layer and the flow rate have minimal effects. However, a low flow rate can result in a significant decrease in the water film temperature. Furthermore, changes in the thermal conductivity of the surface layers also contribute to the anti-icing effect. The simulation analysis of the GCTCS provides valuable guidance for practical engineering in cooler regions where groundwater resources are abundant.

1. Introduction

Winter weather conditions in cold regions often result in frozen road surfaces, leading to increased traffic accidents. Consequently, anti-icing and de-icing measures incur significant costs within road maintenance budgets during this season [1,2,3]. Generally, two approaches are utilized to address snow and ice on pavements: clearance and melting. The clearance method involves mechanically removing snow and ice, while the melting method encompasses chemical and thermal melting techniques [4]. Nonetheless, the use of de-icing salts can cause irreversible damage to roads, impacting their mechanical performance, strength, and fatigue life [5,6,7]. To mitigate the harm inflicted upon structures and the environment, alternative de-icing techniques have been explored, such as incorporating chemical additives like sodium formate and sodium silicate into the asphalt mixture. These additives effectively prevent ice from bonding to the road surface [8]. Furthermore, microwave irradiation is an efficient de-icing method during winter, where the inclusion of additives like steel fibers or magnetite in the mixture enhances the efficiency of microwave heating [9,10,11]. Another environmentally friendly and efficient approach involves utilizing geothermal aquifers to transport heat to the road surface piping system. By circulating water to the road surface, this method effectively reduces icing, surpassing the capabilities of traditional snowmelt systems [12]. Road heating technology is becoming increasingly popular for de-icing roads. This technology involves melting snow on the roads using different methods such as embedding heat circulation pipelines, electric heating wires, carbon fiber heating, or solar heat storage. Compared to traditional methods like mechanical and chemical de-icing, which are inefficient, time-consuming, and harmful to the environment, heating the pavement by embedding pipes in concrete slabs and circulating a heat flux is considered more environmentally friendly and efficient. One promising heat source for these heating systems is groundwater. Groundwater remains at a relatively stable temperature, unaffected by fluctuations in external temperatures and solar radiation. It does not require external heating, making it a renewable and environmentally friendly resource. Groundwater has great potential for melting road snow and is both clean and sustainable. Aquifers of hundreds of or even thousands of meters can provide geothermal water at 150 °C and above. However, for space heating and road snow removal, shallow low-temperature geothermal water is a more appropriate choice [4,13,14].

Compared to ground-level roads, highway bridge surfaces are more prone to freezing due to their elevated position, exposure to high wind speeds, and lower temperatures. Highway bridges are extensively distributed and require higher safety and driving comfort standards compared to regular road surfaces, making the prevention of icing crucial [15,16]. This study focuses on the application of the groundwater circulation temperature control system (GCTCS) for pavements on highway bridges. The GCTCS utilizes a buried pipe-type installation method, where groundwater is pumped into serpentine pipes to maintain a high-speed flow. Through the pipe wall, warm groundwater transmits heat to the bridge surface, effectively de-icing the highway [17]. The GCTCS is particularly suitable for moderately cold areas, such as highway bridges where the annual average daily minimum temperature hovers at around 0 °C. Its application significantly reduces the occurrence of traffic accidents by preventing rain or minor snow accumulation. Groundwater, serving as a clean energy source, is more environmentally friendly compared to traditional snow melting agents [18].

Despite the advantages of the GCTCS, its successful application is influenced by various factors, categorized into environmental conditions, the bridge paving structure, and the design of the GCTCS. Environmental conditions encompass temperature, wind speed, solar radiation, and groundwater temperature. The bridge paving structure includes the properties and thickness of the paving material, as well as the layout of the GCTCS. The GCTCS design parameters consist of pipe size, pipe spacing, and flow rate. Thus, the primary objective of this study was to develop a model of a GCTCS bridge structure to investigate the impact of wind speed, asphalt concrete layer thickness, groundwater temperature, water flow rate, and pipe spacing on the surface water film temperature of the bridge. The aim is to determine the conditions and scenarios in which the GCTCS can be used effectively. This analysis will commence by examining the influence of surface water film temperature under steady-state conditions. Subsequently, real-time temperature changes in the surface water film will be simulated over a 24 h period to replicate cooling conditions that are prevalent in real-world environments. To ensure the reliability of the simulation results, a bridge panel specimen equipped with the GCTCS will be fabricated for laboratory testing and verification. Through an analysis of the impact of external environmental factors and internal structural factors on the successful application of Groundwater Circulation Temperature Control Systems (GCTCS), this study aims to examine the extent to which each factor influences the system. Furthermore, the study seeks to identify the most significant factor and suggest corresponding measures to address the impact of each factor. The ultimate goal is to provide a theoretical foundation and practical guidance for the implementation of these systems in engineering projects.

2. Methodology

2.1. GCTCS System

Figure 1 depicts a schematic diagram of a groundwater circulation temperature control system (GCTCS), which comprises two main components: the pipe network system and the groundwater reservoir system. The GCTCS operates as follows: groundwater is initially pumped into a filter tank. and then, after entering an insulated reservoir, it is pressurized by shallow-buried pumps to move it into the upper section of the pipe network. Subsequently, the water circulates through the pipe network and is returned to the ground through recycling. To ensure system stability, the underground portion of the GCTCS, part of the groundwater reservoir system, is buried in the sub-soil, providing insulation and maintaining a relatively warm underground water temperature. It is crucial to note that if the length of a single pipe network system is excessive, the water flow within the pipeline may not return to the ground in a timely manner, thus affecting the overall efficiency of the GCTCS. Hence, it is more practical to divide the pipe network system into segments [17].

In general, a highway bridge structure equipped with a GCTCS consists of several components, including a T-beam, concrete bridge deck, buried pipe system, concrete pavement layer, and asphalt layer forming the bridge deck pavement (see Figure 2). The pipe system is laid within a cement concrete overlay, assembled and securely fixed to the upper surface of the bridge deck. It is then covered with a thin layer of cement concrete, providing sufficient coverage for the pipe system while ensuring no damage during the compaction process and subsequent traffic opening. The asphalt concrete layer (AC layer) of the bridge deck pavement is typically thinner compared to other layers, ranging from approximately 60 to 100 mm. Other design parameters for the bridge deck pavement can be determined based on the established design methods and engineering experience.

2.2. Physics

The physical model is established with the interfaces that come with COMSOL Multiphysics 5.3a software [19,20], including Heat Transfer in Solids (HT), Heat Transfer in Pipes (HTP), Thin-Film Flow, Shell (TFFS), and Surface-to-Surface Radiation (RAD). TFFS is used to characterize the water film that covers the road surface due to precipitation, and RAD is used to calculate thermal radiation. HTP is employed to simulate heat transfer between the flowing water in each pipe and the external environment, and finally to build a multiphysics coupling simulation with RAD and HT [21]. Therefore, the GCTCS takes into account thermal conduction, thermal convection, and thermal radiation. The following assumptions are primarily made in the model simulation:

• The surface water film is evenly distributed, and the film thickness is taken as 3 mm regardless of evaporation and loss.
• The ambient wind speed is in a stable and continuous condition.
• The temperature of the groundwater flow is constant, i.e., the water temperature at the inlet to the input pipe system is maintained.
• The effect of phase-change submersible heat is ignored for the water film.
• The effect of traffic on the road surface is ignored.

To analyze the energy exchange within the GCTCS inside the bridge, Equation (1) is applied as the governing equation for pipe heat transfer,

$$\rho {AC}_p\frac{\partial T}{\partial t}+\rho A{C}_{p}u{\overrightarrow{e}}_t·{\nabla }_{t}T={\nabla }_t·\left(Ak{\nabla }_{t}T\right)+\frac{1}{2}{f}_D\frac{\rho A}{d_h}{\left|u\right|}^3+Q+Q_{wall}$$

(1)

$$Q_{wall}=\left(hZ{)}_{eff}\right(T_{ext}-T)$$

(2)

$$d_f\rho C_p\left(\frac{\partial T}{\partial t}+\overrightarrow{u}·{\nabla }_{t}T\right)+{\nabla }_t·{\overrightarrow{q}}_f=d_f{Q}_f+q_0$$

(3)

$$Q_f=-d_{f}k{\nabla }_{t}T$$

(4)

where $\rho$ is the liquid density, which can be taken as 1000 kg/m³ for groundwater; A represents the pipe cross-sectional area available for flow; $C_p$ is heat capacity at constant pressure of 4200 J/(kg· K); $u$ is the tangential flow rate because the pipe flow is simplified to a linear form in COMSOL, and therefore $u$ does not have to be calculated by integration; $k$ is the thermal conductivity of 0.57 W/(m· K). The term that differs from the general heat transfer equation is that the second item on the right-hand side of the equation is the heat generated by friction, where $f_D$ is the Darcy friction factor, depending on Reynolds number and pipe properties. $Q_{wall}$ is thermally coupled to an external heat source through a pipe wall, which can be obtained from Equation (2). $(hZ{)}_{eff}$ is a compelling value of the heat transfer coefficient $h$ (W/m·k) times the wall perimeter $Z$ (m) of the pipe. The surface water film is fragile and heat transfers evenly, the water film is thus considered to be a weak thermal approximation. Inside the thin water film layer, the heat equation becomes Equation (3), where $d_f$ is the film thickness, which is taken as 3 mm in this case. The heat source $Q_f$ follows a density-dependent distribution in the layer, while q₀ is the received out-of-plane heat flux.

Then, the heat exchange relationship between GCTCS and external can be established using the following governing equations:

$$q_r=\epsilon \sigma (T_{asp,s}^4-T_{sky}^4)$$

(5)

$$q_{sun}={\alpha }_{s}I·cosi$$

(6)

$$q_c=h_c(T_{asp,s}-T_a)$$

(7)

where $q_r$ is the net intensity of radiation emitted from the pavement surface to the surrounding area [17], according to the Stefan-Boltzmann law, and $q_r$ depends on the fourth power of the temperature in Equation (5). $\epsilon$ is the road surface emissivity, 0.96; $\sigma$ is the Stefan-Boltzmann constant, 5.67 × 10⁻⁸ W/(m²·K⁴); and $T_{sky}$ is the sky radiation temperature. $q_{sun}$ is the energy absorbed from direct solar radiation in Equation (6); ${\alpha }_s$ is the sunlight radiation absorptivity of the asphalt mixture, 0.90 [22]; $I$ is the solar radiation impacting the Earth’s surface, set to 200 W/m² under the winter conditions pertinent to this study; $i$ is the angle between the normal and surface and the direction of radiation. The convectional thermal equation for the GCTCS bridge and the external environment is given by Equation (7), where $h_c$ is the convection heat transfer coefficient, W/(m²·K), $h_c$ = 5.6 + 4.0 v_w, v_w is the wind speed in m/s, and $T_a$ is the air temperature, °C.

The boundary conditions for the GCTCS bridge are shown in Figure 3. As the position of the sun changes over time and the radiation angle changes at all times, the surface of the pavement and the edges on both sides of the bridge are affected by solar radiation; therefore, they can be defined as exposure boundaries with thermal radiation effects imposed by the surrounding environment. For the boundary conditions on the lower side, there are many studies on temperature gradient distribution in beams [2,23], but this is not the main research focus here. Considering the particularity of the three compartments under the beam and analyzing the heat transfer characteristics, the boundary condition beneath the beam is taken into account in the following ways [24,25].

• The bridge deck slab connects to the infrastructure through the abutment, and there is thus little thermal conduction.
• The longitudinal axis of the bridge, the bottom area of which is sheltered by the lateral T-beams, is not exposed to solar radiation.
• Wind slows down in the cavity, which reduces heat convection loss (Figure 4 shows the wind speed distribution across the T-beam structure under the action of transverse wind, and the wind speed in the cavity therein is decreased).

In addition, the condition of steady-state models arises at night, i.e., no solar radiation effect is considered, while the transient model considers the changes in solar radiation over time.

2.3. Settings

Table 1. Material and structural parameters for each component in the GCTCS bridge.

Name and Parameters	Materials
	AC Layer	Concrete Layer	Groundwater	Water Film
	Asphalt Mixture	Cement Concrete	Liquid Water	Liquid Water
Density (kg/m3)	2400	2300 *	1000	1000
Heat capacity kJ/(kg·K)	1.0	0.88 *	4.2 *	4.2 *
Thermal conductivity W/(m·K)	1.3	1.8 *	0.57 *	0.56 *
The ratio of specific heats			1.0 *	1.0 *
	Structures
Thickness (mm)	60~100	30		3
Length and width of individual module (m)	5.5 (L) 7.5 (W)	5.5 (L) 7.5 (W)		5.5 (L) 7.5 (W)
Pipe spacing (mm)			100~300
Water flow rate (L/min)			25~125

Note: (*) denotes a COMSOL built-in material parameter, others refer to [17].

and

Table 2. Environmental conditions.

Ambient Temperature	Wind Speed	Groundwater Temperature	Bridge Location
−3 °C	1.6~10.8 m/s	8~20 °C	29°15′16″ N, 115°44′08″ E

present the material, structure, and environmental parameters of the GCTCS system. The GCTCS is divided into sections, each with its own water inlet and outlet (see Figure 1). To mitigate boundary effects, the analysis focuses on the mid-section of the GCTCS. The bridge’s geographical location is in Poyang County, Jiangxi Province, China, where winter temperatures can plummet to as low as −1 °C to −4 °C. The wind level corresponds to the lower wind limit of 2 to 6, as specified by the international standard Beaufort Scale [26]. The simplified model for simulation is displayed in Figure 5. It was divided into 148,377 grids, consisting of 24,840 grid vertices, 113,837 tetrahedral mesh elements, 34,540 triangular mesh elements, 6910 edge elements, and 396 vertex elements.

2.4. Experiment

In this research, a 1:1 scale specimen of a bridge deck was employed in the laboratory experiment to validate the numerical model. The specimen consisted of two layers of asphalt concrete and two layers of cement concrete, namely the concrete levelling course and the bridge deck slab (see Figure 6a). The two asphalt concrete panels produced were a of types AC-13 and AC-20, with thicknesses of 4 cm and 6 cm, respectively. Their gradation curves are shown in Figure 7 [27]. Before pouring the concrete, the pipes with five U-shaped turns were carefully positioned within the concrete levelling course (see Figure 6b). The spacing between embedded pipes was 20 cm, and they were situated beneath the surface of the top layer. Additionally, a protective layer, approximately 3 cm thick, was applied along the edges of the specimen to minimize boundary heat dissipation. The specimen was equipped with an inlet and an outlet to connect an external pump, enabling the establishment of a water circulation loop (see Figure 6c).

The experimental setup was conducted in a spacious test chamber that offered precise temperature control, allowing for high and low temperature conditions. To minimize heat dissipation, the side and bottom surfaces of the specimen were wrapped with an insulating material (see Figure 6c). Furthermore, the specimen surface was fully submerged in water to simulate a wet pavement condition with a water film, and three temperature sensors were strategically placed on the specimen surface to capture real-time temperature variations based on the test scenarios (see Figure 6d). However, it is important to note that, due to experimental limitations, controlling certain factors in the laboratory setting proved challenging. Consequently, we focused solely on water temperature and the water flow rate in our tests, as these factors are relatively easier to adjust and demonstrate a strong correlation. The temperature and flow rate of the water within the pipes were controlled by a large thermostatic water tank equipped with a pump.

3. Results

3.1. Steady-State

The steady-state results for the average temperature, maximum temperature, and minimum temperatures of the water film on the surface were analyzed based on five single-factor variables. The experiments were conducted under the same conditions, and the findings are presented in Figure 8.

• Wind speed directly impacts the thermal convection of the surface, leading to a reduction in the temperature of the water film as wind speed increases. However, the cooling effect slows down. The simulation result demonstrates the significant influence of wind speed on the water film temperature (see Figure 8a). For instance, as the wind speed increased from 1.6 m/s to 10.6 m/s, the average water film temperature decreased by 5.6 °C in the simulation. When the wind speed exceeded 10 m/s, the minimum temperature of the water film dropped below 0 °C;
• The variation in the asphalt concrete (AC) layer’s thickness determines the distance between the pipe system and the surface water film, directly affecting the heat transfer. The simulation result indicates that increasing the AC layer thickness leads to a gradual decrease in the water film temperature, along with a reduction in the difference between the maximum and minimum temperatures (see Figure 8b).
• Groundwater temperature plays a crucial role in determining the temperature difference between the interior and exterior of the pipe, thereby affecting heat transfer. Both the simulation and experiment reveal a linear increase in the surface water film temperature with respect to groundwater temperature. When the groundwater temperature reached approximately 13 °C, the minimum temperature remained above the freezing point of water (see Figure 8c).
• The water flow rate determines whether warm water can enter the pipe system in a timely manner for heat release. Increasing the flow rate from 25 L/min to 125 L/min resulted in a relative growth rate of 400%. However, the water film temperature shows a slow increase, with the average temperature rising from 1.65 °C to 2.15 °C in the simulation (see Figure 8d). The experimental results show slightly lower water film temperatures compared to the simulated values, but the overall trend remains the same.
• Pipe spacing governs the temperature field distribution and determines the sufficient flow time for groundwater within the pipe network system. It serves as a comprehensive indicator of system behavior. With an increase in pipe spacing, all four temperature indicators decrease to varying extents, with the minimum temperature exhibiting the most significant reduction, followed by the average temperatures. At a pipe spacing of 100 mm, the temperature difference is only 0.47 °C, whereas at a spacing of 300 mm, the temperature difference rises to 2.65 °C, and the minimum temperature falls below freezing (see Figure 8e).

3.2. Transient Effects

In practice, cooling is typically a process of decreasing temperatures until they eventually stabilize. Therefore, it is essential to calculate the transient state of the GCTCS bridge continuously as the temperature gradually decreases over time. To simulate the cooling of the environment, the real temperature changes over 24 h were taken into account. Specifically, the simulation of ambient temperature began at 3 p.m. on the first day, and a continuous cooling process was modeled. The minimum temperature of −4 °C was reached after 12 h and remained relatively stable until 3 p.m. on the following day (see Figure 9a).

With regard to solar irradiance, the exposed boundary received no solar radiation between 5 p.m. on the first day and 8 a.m. the next day (see Figure 9b). As it is challenging to precisely replicate solar radiation in laboratory experiments, a simulation was adopted in this study to obtain the changes in water film temperature on the bridge deck under the combined effects of ambient temperature and radiation.

The variables considered were wind speed, AC layer thickness, groundwater temperature, pipe water flow rate, and pipe spacing, each at four levels. Following the design of the orthogonal analysis scheme based on the five-factor four-level orthogonal table L16(4⁵), the appraisal index comprised the average temperature, minimum temperature, and lower limit within 24 h on the surface of the bridge pavement deck (see

Table 3. Orthogonal experimental design for simulation.

Group	Wind Speed (m/s)	AC Layer Thickness (mm)	Groundwater Temperature (°C)	Flow Rate (L/min)	Pipe Spacing (mm)
A	3.4	70	11	50	100
B	3.4	80	14	75	150
C	3.4	90	17	100	200
D	3.4	100	20	125	250
E	5.5	70	14	100	250
F	5.5	80	11	125	200
G	5.5	90	20	50	150
H	5.5	100	17	75	100
I	8.0	70	17	125	150
J	8.0	80	20	100	100
K	8.0	90	11	75	250
L	8.0	100	14	50	200
M	10.8	70	20	75	200
N	10.8	80	17	50	250
O	10.8	90	14	125	100
P	10.8	100	11	100	150

).

Figure 10 and Figure 11 show the average temperature and minimum temperature of the water film during the 24 h cooling process. The average temperature of the water film rose, and then decreased within the next 12 h. The early temperature rise is due to the ambient temperature cooling range being small; in addition, heat in the pipeline heat flow accumulates in the water film. After reaching the maximum temperature, due to the reduction of ambient temperature and the disappearance of solar radiation, a gradual decrease in the temperature of the water film occurred. It reached the stable state of a minimum temperature about 12–17 h into the simulation. After 17 h, the ambient temperature did not change significantly, and the exposure boundary temperature increased due to solar irradiation. The solar radiation input weakened at about 22 h, and the temperature gradually decreased thereafter. In Figure 11, the minimum temperature of the water film (after 20 h) can be seen to significantly increase because concrete guardrails obscured the sunlight on both sides before this time. With the decrease in the solar radiation angle ($i$ in Equation (6)), the temperature of the water film gradually reaches a local maximum. In general, the change of water film temperature conforms to the change of ambient temperature and solar irradiance.

Table 4. Significance analysis: orthogonal experimental design.

Indicator	Factor	Analysis of Indicators
		$\overline{{\mathit{K}}_1}$	$\overline{{\mathit{K}}_2}$	$\overline{{\mathit{K}}_3}$	$\overline{{\mathit{K}}_4}$	$\mathit{R}$
The lower limit of the average temperature of the water film in 24 h (°C)	Wind speed	2.541	1.168	0.395	−0.477	3.018
	AC layer thickness	1.166	1.000	0.855	0.606	0.560
	Groundwater temperature	−0.117	0.555	1.272	1.917	2.034
	Flow rate	0.716	0.868	1.036	1.007	0.320
	Pipe spacing	1.283	1.174	0.696	0.474	0.809
The lower limit of the minimum temperature of the water film in 24 h (°C)	Wind speed	1.797	0.486	−0.224	−1.093	2.890
	AC layer thickness	0.339	0.393	0.223	0.012	0.381
	Groundwater temperature	−0.651	−0.109	0.584	1.144	1.795
	Flow rate	0.029	0.222	0.387	0.33	0.358
	Pipe spacing	0.956	0.589	−0.031	−0.547	1.503

presents the significant components derived from the selected orthogonal design, in which factors with higher statistical significance exert a more substantial influence on the appraisal indicator. The analysis reveals that wind speed has the most significant impact. Following wind speed, groundwater temperature demonstrates the second-highest importance, while the significance of pipe spacing, surface thickness, and flow rate decreases in that order. This finding aligns with the conclusions drawn under steady-state conditions. Notably, the significance of pipe spacing increases in later stages.

Table 5. Analysis of variance: orthogonal experimental design.

Indicator	Factor	Sums of Squares (SS)	Degrees of Freedom (DF)	F-Ration (F)	Significance
The lower limit of the average temperature of the water film in 24 h (°C)	Wind speed	19.654	3	76.178	*
	AC layer thickness	0.675	3	2.616
	Groundwater temperature	9.3	3	36.047	*
	Flow rate	0.258	3	1
	Pipe spacing	1.781	3	6.903
	Error	0.258	3
The lower limit of the minimum temperature of the water film in 24 h (°C)	Wind speed	17.906	3	59.886	*
	AC layer thickness	0.341	3	1.14
	Groundwater temperature	7.403	3	24.759	*
	Flow rate	0.299	3	1
	Pipe spacing	5.314	3	17.773	*
	Error	0.299	3

displays the results of the variance analysis of the orthogonal designs at a significance level of 0.05. No error column is included, and the flow rate of the minimum sums of squares (SS) serves as the error term. The conclusion drawn from the analysis is that wind speed and groundwater temperature are significant factors influencing the average temperature index of the water film. Wind speed, groundwater temperature, and pipe spacing are crucial for determining the minimum temperature index of the water film. The greater significance among these factors has been marked by * in the Table 5.

4. Discussion

4.1. Wind Speed

The impact of wind speed on the temperature of the water film is a critical consideration in GCTCS design. In the current study, it was observed that when the wind speed surpasses 10 m/s, the water film freezes rapidly, as depicted in Figure 8a. Analyzing the proportion of heat convection to heat loss in each simulation revealed that as the wind speed increases, the proportion of heat loss also increases significantly, as shown in Figure 12. This increase in the proportion of heat loss becomes less pronounced once the wind speed exceeds 3 m/s and can account for over 90% of the observed variation. Additionally, considering the temperature characteristics in cold regions, it is essential to acknowledge that higher bridge deck altitudes lead to increased wind speeds. Consequently, greater attention should be given to the influence of wind speed when designing a GCTCS.

The wind speed, on the other hand, is a geographical characteristic of a region, and the model assumes a constant and continuous wind speed. However, predicting the wind speed and direction in the environment can be complex. In terms of highway bridge wind speed calculations, the most critical scenario is the occurrence of strong and sustained gusts, which can lead to a brief freezing of the pavement surface. Therefore, such gusts should be taken into account during the design of GCTCS. To address this, wind speed measurements should be conducted at the bridge location, and a location scheme that minimizes wind speed is preferred. Additionally, if necessary, the installation of windbreaks on both sides of the bridge can be considered. The use of windbreaks to reduce wind speed and enhance anti-icing capabilities is a novel research idea for highway bridges, although it has already been extensively applied in railway bridge engineering [14,28,29,30].

4.2. Groundwater Temperature

The influence of groundwater temperature on the temperature of the water film is also a significant factor, and a linear relationship between them can be observed (see Figure 8c).

Table 6. Temperature of water flow in GCTCS pipes.

Model (Wind Speed/AC Layer Thickness/Groundwater Temperature/Flow Rate/Pipe Spacing)	Inlet (°C)	Outlet (°C)	Difference (°C)
5.5 m/s, 90 mm, 14 °C, 75 L/min, 200 mm	14	13.1	0.9
5.5 m/s, 90 mm, 17 °C, 75 L/min, 200 mm	17	15.9	1.1
5.5 m/s, 90 mm, 20 °C, 75 L/min, 200 mm	20	18.7	1.3
5.5 m/s, 60 mm, 17 °C, 75 L/min, 200 mm	17	15.7	1.3
10.8 m/s, 90 mm, 17 °C, 75 L/min, 200 mm	17	15.8	1.2
5.5 m/s, 90 mm, 17 °C, 75 L/min, 150 mm	17	15.9	1.1

presents the inlet and outlet water temperatures obtained from simulations involving the six aforementioned models, along with the temperature difference between them. In the table, groups A, B, and C represent variations in groundwater temperature, group D involves a reduction in the thickness of the AC layer, group E corresponds to an increase in wind speed, and group F pertains to a decrease in pipe spacing. The water flow rate was kept constant across all six models, and the temperature difference between the inlet and outlet indicates the amount of heat conducted by the fluid flow through the pipe. A higher temperature difference signifies a greater amount of heat dissipation.

For GCTCS bridges, it is necessary to explore suitable groundwater extraction locations, which requires a complete geological survey process. Groundwater temperatures are related to local geographical conditions, environmental factors, climate change, and depth [31,32,33,34]. Before designing the GCTCS system, a survey of the groundwater temperature is essential (especially in winter). The temperature of the groundwater in the area around Poyang Lake, for instance, in Nanchang City, Jiangxi Province, is between 17 °C and 20 °C all year round, the depth to groundwater is 5 to 15 m, and the range of water level change is about 2 to 2.5 m [35]. If the groundwater temperature is below the demand threshold, it is necessary to consider heating the groundwater. If the local winter solar irradiance is sufficient, the use of a solar heat collection system is preferred [36,37]. Heating groundwater with solar energy is more effective than direct heating on the pavement surface, as warm groundwater can be stored in the filter tank or reservoir for long periods for use in the event of sudden temperature drops [38,39,40].

4.3. Thermal Conductivity

Thermal conductivity is not explicitly addressed as one of the factors in this paper; however, it is important to acknowledge its significance in the discussion. Cement and asphalt concrete, the materials used in the bridge structure, have relatively low thermal conductivity [41,42]. The temperature distribution across the model’s cross-section (see Figure 13) reveals distinct patterns. In the middle portion of the model, temperatures ranged from 14 °C to 16 °C (indicated by the orange color, labeled as A). Conversely, the temperature along the exposed boundaries of the bridge remained below 0 °C (represented by the blue color, labeled as C). Notably, the area spanning from the bottom of the cement concrete layer to the T-beam belly plate position exhibited significant and unfavorable temperature changes (illustrated in green, labeled as B).

The relatively low thermal conductivity of the concrete underneath and around the bridge panel plays a crucial role in preserving the temperature and slowing down heat loss from the system, making it advantageous for implementing a GCTCS. However, it is essential to note that the AC layer and concrete layer, where the pipe system is located, should be designed to be as large as possible to enhance thermal conductivity. This design choice facilitates a more efficient transfer of heat from the flowing water inside the pipes to the surface water film. To further enhance the thermal conductivity of asphalt mixtures, incorporating high-thermal-conductivity materials like graphite or carbon fiber is recommended. Studies have shown that the addition of graphite, milled carbon fiber, or chopped carbon fiber to a conductive filler can improve the thermal conductivity of asphalt mixtures to around 2.0 to 2.3 W/(m·K) [43]. Moreover, the incorporation of graphite powder filler in asphalt mixtures can lead to up to a 43% improvement in thermal conductivity [44]. These strategies can contribute significantly to the efficiency and performance of the GCTCS system.

By enhancing the thermal conductivity of the AC layer and concrete layer, the average temperature of the water film experiences varying degrees of increase, as shown in

Table 7. Variations in thermal conductivity and film temperature.

Model (Wind Speed/AC Layer Thickness/Groundwater Temperature/Flow Rate/Pipe Spacing)	AC Layer Thermal Conductivity W/(m·K)	Concrete Layer Thermal Conductivity W/(m·K)	The Average Temperature of Water Film (°C)
A. 10.8 m/s, 90 mm, 11 °C, 75 L/min, 200 mm	1.3	1.6	−0.70
	2.5	2.2	0.50
B. 5.5 m/s, 90 mm, 17 °C, 75 L/min, 200 mm	1.3	1.6	1.92
	2.5	2.2	4.12
C. 3.4 m/s, 90 mm, 20 °C, 75 L/min, 200 mm	1.3	1.6	4.29
	2.5	2.2	7.11

. In particularly adverse conditions with a wind speed of 10.8 m/s and a groundwater temperature of 11 °C, the increase in thermal conductivity results in a 1.2 °C rise in the average water film temperature. This finding indicates that the implementation of a GCTCS can be considered to be feasible in regions characterized by high wind speeds and low groundwater temperatures.

Similarly, for conditions involving moderate wind speeds and groundwater temperatures, the impact of increased thermal conductivity on temperature is also evident. Therefore, it is recommended to use a cement concrete material with twice the thermal conductivity of cement mortar for the covering layer. Additionally, the paving layer for the pipe system should be constructed using cement concrete instead of cement mortar. These measures contribute to a more effective utilization of the GCTCS system, especially in regions where moderate wind speeds and groundwater temperatures prevail.

4.4. Pipe Spacing

Pipe spacing is a controllable factor that exerts a considerable influence on the minimum temperature of the water film. It is evident that the lowest temperature of the water film occurs at the center, farthest from the two pipes (see Figure 14). By examining the cross-sectional temperature field distribution and isothermal lines, it becomes apparent that smaller pipe spacing (i.e., denser arrangement) results in the isothermal contours of the AC layer and water film being closer to the horizontal line, leading to a more uniform temperature distribution within the pavement along the direction of traffic flow.

However, it is crucial to consider the cost implications of the pipeline system, as it constitutes a significant portion of the overall cost of the GCTCS. Therefore, determining the appropriate pipe spacing should take into account local conditions. In areas where the water film temperature may approach the freezing point of water, it is advisable to avoid excessively large pipe spacings to prevent the formation of intermittent icing regions. Conversely, economic factors can be considered to expand pipeline spacings as is deemed appropriate. For instance, if a serpentine arrangement is adopted with a pipe outer diameter of 26.7 mm, it is recommended to keep the pipe spacing within 300 mm. By striking the right balance between preventing icing and optimizing cost efficiency, an effective GCTCS design can be achieved based on local conditions and requirements.

4.5. AC Layer Thickness and Flow Rate

For both the steady-state and transient simulations, the thickness of the AC layer and water flow rate had a small effect on the temperature of the water film. Figure 15 shows the temperature of a wheel path in the water film layer along the direction of the traffic flow (parallel to the mid-line of the bridge, offset by 500 mm). The thinner AC layer reduces the distance from the pipe to the surface water film and accelerates the heat transfer, but the distribution of the surface temperature may not be sufficiently uniform. However, the thickness of the AC layer on the bridge deck will generally not be less than 60 mm; in practical applications, a reasonable thickness of the surface layer can be obtained according to structural pavement design calculations and the performance of the asphalt mixture.

The flow rate is also an indicator that is under human control, and the best GCTCS’ mode of operation is a flow-rate adjustable set-up. Although the flow rate has little effect on the temperature of the water film, the water film temperature will also be significantly reduced (see Figure 6d), and if it is too low (less than 25 L/min), it may freeze in the pipe. The general water pipe flow rate should not be higher than 2.0 m/s, or 50 L/min. On the contrary, a high flow rate will significantly increase the water pressure on the pipe wall, which is not recommended.

5. Conclusions

The groundwater circulation temperature control system (GCTCS) is a viable solution for preventing the freezing of wet pavement surfaces during winter in slightly cold areas. In this study, the GCTCS was simulated using COMSOL to investigate the temperature changes of the surface water film under steady-state and transient conditions. Several influencing factors were examined, including wind speed, AC layer thickness, groundwater temperature, pipe water flow rate, and pipe spacing. The following conclusions can be drawn:

• Wind speed emerges as the most influential factor among the five considered. As wind speed increases, the temperature of the water film experiences a sharp decline. This cooling effect is particularly pronounced at low wind speeds. When the wind speed exceeds 10 m/s, the water film temperature can rapidly drop below the freezing point, with convective heat loss accounting for over 90% of the overall heat loss.
• Groundwater temperature ranks second only to wind speed in terms of its impact on the heat transfer process. A direct linear relationship exists between groundwater temperature and the steady temperature of the water film, with higher groundwater temperatures resulting in higher water film temperatures.
• Pipe spacing noticeably affects the minimum temperature of the surface water film, while the thickness of the AC layer significantly influences the uniformity of the temperature distribution within the water film. Conversely, the flow rate exhibits minimal influence on these factors.
• Enhancing the thermal conductivity of the AC layer and concrete layer proves beneficial in elevating the water film temperature.

In summary, utilizing groundwater as a clean energy source for bridge anti-icing systems offers an environmentally friendly and efficient approach. The modeling and simulation of the GCTCS provide valuable guidance for its implementation in engineering practice. To optimize its effectiveness, GCTCS bridge locations should be situated far from high-wind-speed areas. Additionally, measures such as windbreaks can be employed to mitigate convective heat loss. While groundwater temperature, like wind speed, remains beyond our control, seeking higher-temperature groundwater sources and insulating storage tanks are crucial considerations. The cost of the pipeline plays a significant role in determining the feasibility of the installation and use of a GCTCS. For a GCTCS employing a 26.7 mm pipe outer diameter, it is recommended to maintain a pipe spacing not exceeding 300 mm. The thickness of the AC layer can be designed based on structural strength and durability requirements. Alternatively, enhancing the thermal conductivity of the AC layer and concrete layer provides an alternative approach.

This study oversimplifies some environmental factors, such as assuming a constant wind speed and water flow rate within the pipes, and neglects the impact of traffic and rainfall. Future research should investigate these aspects in greater depth to improve the experimental conclusions. Groundwater, as a clean and sustainable energy source, shows great potential for controlling pavement temperature and de-icing. It offers significant advantages compared to traditional heat sources. Regions with ample groundwater resources, particularly those with low temperatures, should prioritize research on using groundwater for pipeline heating systems. This approach aligns perfectly with the concept of sustainable road development.