Experimental Study on Ice Melting of Bridge Tower Crossbeams under Different Conditions

: The bridge tower beams of long-span bridges freeze during winter. The accumulated ice on the beam ﬂakes off and falls into the bridge deck from a high altitude with temperature rise, threatening trafﬁc safety. To solve this problem, an electric heating method is used to melt the snow and ice on bridge tower beams. A design scheme based on embedding carbon ﬁber heating wires in a bridge tower beam is proposed in which electrical energy is converted into heat energy to melt the snow and ice on the beam. In this paper, a beam icing test and beam de-icing test are carried out through a laboratory model test. The model test results show that under the conditions of a wind speed of 4.5 m/s and an ambient temperature of − 15 ◦ C, the average internal temperature of the beam after heating stabilization is 1.37 ◦ C; this is higher than the average critical temperature of 1.31 ◦ C. This temperature is sufﬁcient to melt the ice and proves the practicability of this method. The de-icing method of embedded carbon ﬁber heating is simple, efﬁcient, and environmentally friendly. This can provide a reliable reference for the practical application of de-icing a bridge beam.


Introduction
Roads and bridges are an important part of transportation infrastructure and play a vital role in social and economic development. During the cold winter months, most traffic accidents are caused by icy and snowy roads [1]. However, traffic accidents do not only occur on roads. For example, large-span bridge tower crossbeams are located in a harsher environment prone to icing. When the temperature is low, fog, rain, and snow condense on the surface of bridge tower crossbeams to form an ice layer, which thickens. As the temperature rises, the ice layer melts at the contact surface with the crossbeam and ties and easily falls off into the lane below under the action of gravity, resulting in a large area of falling ice [2]. Due to the large icing surface area of the crossbeam, the water is easily attached to form a large mass of ice, creating dangerous conditions when the ice layer falls off. According to statistics, there are many cable-stayed bridges or suspension bridges with beams which produce falling ice, resulting in a significant safety hazard for bridge traffic [3]. On 8 January 2018, an alarming scene occurred on the Erqi Yangtze River Bridge in Wuhan. Vehicles were driving on the bridge, and ice suddenly fell from the sky and the traffic on the bridge was brought to a standstill. On 31 January 2015, more than 20 vehicles were hit by icy snowfall on the Erqi Bridge in Wuhan. According to statistics, many Yangtze River bridges, such as the Wuhan Baishazhou Bridge, the Second Yangtze River Bridge, Yingwuzhou Bridge, and Junshan Bridge, have experienced ice debris, which constitutes a significant hidden hazard when driving on the bridge deck. Thus, it is necessary to take targeted measures to melt the ice of the beams, with the surface of the beams removing icicles and snow, so as to ensure the safety of bridge traffic in winter. Mengjie Bao from Chang'an University [4] conducted research on built-in carbon fiber ice-melting pavement. She used the method of built-in carbon fiber heating wires in this experiment and embedded the carbon fiber heating wires during the process of pouring pavement concrete. Subsequently, she analyzed the external environment's temperature and wind speed. Multiple factors were found to have an effect, such as laying spacing of the fibers affecting the surface temperature of the built-in carbon fiber ice-melting pavement. The built-in carbon fiber bridge pylon beam used in this study is not exactly the same as the heat transfer inside the concrete of the built-in carbon fiber pavement, and thus this study is different from others in the literature because the research objects are different, and it needs to be studied separately.
At present, domestic and foreign research on bridge de-icing is mainly focused on the bridge deck and rarely applied to the bridge tower beams. This is due to the great height of the bridge tower beams, for which bridge deck salt or mechanical de-icing and other conventional methods are not feasible. Currently, the active bridge deck de-icing methods are conductive concrete, geothermal methods, electric heating methods, and various other ways. The conductive concrete method constitutes mixing steel fiber [5], carbon fiber [6,7], graphene [8], and other conductive materials into the concrete so that the concrete has a certain conductivity. At the same time, due to its own resistance, the concrete can generate a certain amount of heat to melt the ice and snow after being energized. However, over time and with frequency of use, the steel fiber will rust, resulting in a decrease in electrical conductivity [9], which will consume more electrical energy in the heating process. In addition, the method has a limited effect when applied to bridge tower crossbeams. The geothermal method [10][11][12][13] involves the installation of heat transfer tubes inside the road concrete, using liquid as the circulating medium, to transfer heat energy from the ground to the road surface for de-icing [14,15]. However, the equipment and methods used to exploit geothermal energy are complex; furthermore, the bridge tower beams are high, and the heat loss from the heat transfer pipes leads to a significant compromise in ice melting [16]. The electric heating method is a simple, efficient, and environmentally friendly de-icing method that uses electricity as the energy source and a carbon fiber heating line [17] as the heat generator to convert electrical energy into thermal energy. This method transmits heat to the surface of the object through thermal conduction within the structural layer, and then melts snow and ice through the exchange of sensible and latent heat between the surface of the object and the snow and ice [18,19]. It is theoretically feasible to consider the electric heating method to increase the temperature of the beam structure to achieve a melting effect on snow and ice.
Researchers have conducted a series of studies on the application of carbon fiber heating lines in snow and ice melting under different conditions. Yan Tan et al. [20] determined that carbon fiber heating lines can effectively prevent bridge deck icing through temperature rise tests on bridge decks and found that optimized layouts can improve the efficiency and overall uniformity of bridge deck ice melting. Chunming Li and Dengchun Zhang et al. [21,22] conducted a temperature rise test on a bridge deck. It was concluded that the temperature difference between different locations was within 3 • C, with good temperature uniformity, when the spacing of carbon fiber heating lines was 10 cm. Mohammed et al. [23] embedded three different types of carbon fibers into concrete road samples. The carbon fiber electric heating method can effectively solve the snow and ice problem, which not only has the effect of energy saving and environmental protection, but also can be used for a long period of time at low cost and with fast results. Hongming Zhao et al. [24] embedded carbon fiber heating lines in concrete pavement, which yielded excellent field test results for concrete slabs with appropriate input power and snow accumulation on slabs with thicknesses of 10, 40, and 70 mm at 0.75, 2.5, and 3.5 h, respectively. The burial depth and arrangement density of carbon fiber heating lines [25] affect the heat transfer from the inside, whereas environmental factors such as ambient temperature and wind affect it from the outside [26,27]. Therefore, the influence of internal and external factors on the heat transfer of the carbon fiber heating line should be fully considered, and it is necessary to systematically analyze the electric heating method for melting snow and ice with respect to the structural characteristics of the bridge tower beams and the external environmental factors.
In order to obtain the internal and surface temperature field distribution of the bridge tower crossbeam under the built-in carbon fiber heating condition, the bridge tower crossbeam model test was carried out at a constant temperature in a large test chamber under different ambient temperature and wind speed conditions. The aim was to study the influence of ambient temperature and wind speed on the internal and surface temperature fields of the bridge tower beam model, to verify the feasibility of the electric heating technology applied to the bridge tower beam for melting snow and ice, and to provide a reliable reference for the actual engineering application.

Experimental Design
In accordance with the main tower crossbeam structure of the actual bridge, a C50 concrete crossbeam model was constructed at an outdoor site. The size of the model was 2 m × 0.4 m × 0.2 m, and a layer of carbon fiber hot wire and temperature measurement points were laid in plane layer A at the buried depth of 3 cm on the side surface of the concrete crossbeam model and in plane layer B at the buried depth of 3 cm on the bottom surface of the concrete crossbeam model. In the test, a 14 m long 48 K carbon fiber hot wire was used. The hot wire entered plane layer A from the left end of the model, and bent to the inside at 10 cm from the bottom of the beam model to reach plane layer B. The hot wire came out from the left end of the model through plane layer B, as shown in Figure 1. The detailed design parameters of the model are shown in Table 1.  In order to accurately measure the temperature field changes inside the beam model during the test, the K-type temperature measurement line was arranged on the carbon fiber heating line according to the arrangement form of the carbon fiber heating line and the structural layer distribution of the specimen, as shown in Figure 2. Figure 2a is divided into three temperature measurement groups from top to bottom: A, B, and C. Each group has three temperature measurement points, marked (from left to right) A 1 , A 2 , and A 3 for group A; B 1 , B 2 , and B 3 for group B; and C 1 , C 2 , and C 3 for group C, totaling nine temperature measurement points. Figure 2b shows group D, with three temperature measurement points, marked (from left to right) D 1 , D 2 , and D 3 , totaling three temperature measurement points.

Model Making
(1) In order to prevent the carbon fiber hot wire being buried in a shifted position, the carbon fiber hot wire was pre-buried in the concrete beam model attached to the glass fiber grid. The size of the glass fiber grid was 2 m × 0.6 m. The glass fiber grid was used to attach the carbon fiber hot wire, but also to enhance the strength of the concrete beam, to extend the life of the beam, and to prevent the formation of beam reflection cracks. Since the selected carbon fiber hot wire is hard when it is not energized at room temperature, it was attached in the upper part of the glass fiber grating by softening the hot wire with electricity. In order to avoid the deflection of the carbon fiber heating line when the concrete is poured, nylon ties were used to attach it. The finished carbon fiber heating arrangement is shown in Figure 3. A 2 m × 0.4 m × 0.2 m wooden board mold was made, and a support frame was set up at the upper end to prevent the mold from being squeezed. The temperature measurement points were divided into four groups: A, B, C, and D. Three groups, A, B, and C, were arranged on plane layer A, with a total of 9 temperature measurement points, and group D was arranged on plane layer B, with a total of 3 temperature measurement points. In order to avoid the deviation of temperature measurement points when the concrete was poured, nylon ties were used to attach each temperature measurement point on the reinforcement cage. The layout of temperature measurement points is shown in Figure 4. (2) In accordance with the tower crossbeam structure of the actual bridge, C50 strength concrete was used in the test to make the crossbeam model. With the convenience of the test taken into consideration, the concrete was poured in layers with the pre-placement of the carbon fiber heating lines attached to the glass fiber grids in the crossbeam molds, and the concrete pouring, pounding, and smoothing were undertaken as shown in Figure 5. (3) After vibration of the concrete beam model was completed, the beam model was covered with a film. The purpose of the film covering was to separate the moisture from the concrete surface and the atmosphere and to reduce the evaporation of water. Film covering makes water react fully with cement in concrete and reduces early cracks in concrete caused by internal stresses due to insufficient water during hydration. The film was added as a cover after the concrete surface had finished vibrating and absorbing water. After the water has been absorbed by the concrete, the film is better able to cover the concrete surface. After 12 h of film coverage, the film was opened and water removed to ensure that the concrete remained moist, and the curing time was 28 days. The finished concrete beam model is shown in Figure 6.

Test Systems
The following auxiliary equipment and devices were required to conduct this model test: a sprinkler system to simulate rainfall, a wind speed simulation system to provide wind speed, a monitoring system to observe and record test conditions, and a data acquisition system to monitor and record temperature changes.
(1) Spraying system The spraying system is composed of a chiller and a nozzle. The chiller is an air-cooled low-temperature chiller. Its main function is to reduce the temperature of room temperature water to 0.5~1.5 • C, to speed up the icing rate to resemble the actual situation. The nozzle is an air atomization nozzle with a 0.5 mm water outlet diameter, produced by Jelly Spray Technology. Its main function is to reduce the particle size of raindrops to a certain extent, thus speeding up the condensation rate of raindrops and making icing easier. There are 3 nozzles, each with a spacing of 0.5 m and a height of 1.7 m from the beam.
(2) Wind speed simulation system The air supply system consists of a fan and a regulator. The fan is an axial fan with a power of 2 KW, rated voltage of 380 V, air volume of 18,700 m 3 /h, air supply diameter of 610 mm, and maximum wind speed of 8 m/s. Its main function is to provide wind speed. The regulator is a three-phase regulator. By connecting it in series with the axial fan, the voltage of the fan can be changed, so as to change the wind speed, using an anemometer to determine the wind speed.
(3) Surveillance system The monitoring system consists of a camera and video recorder. The camera resolution can reach 2560 (horizontal) × 1440 (vertical) pixels, the video frame rate can reach 30 fps, the highest access to the video recorder is 400 W pixels, and the output resolution can reach 1080 P. Its main function is to monitor and record the test conditions in real time, so as to better analyze the test process.
(4) Data acquisition system The data acquisition system consists of a thermocouple line and a data collector. The thermocouple line is a K-type line, and the temperature measurement range is from −50 • C to +160 • C. The data collector has 20 channels with a maximum voltage of 120 V and a maximum current of 20 mA. By connecting this with the thermocouple line, the temperature change can be monitored and recorded in real time.

Constant Temperature test Chamber with Carbon Fiber Heating Line
This test was used to simulate equipment used in a low-temperature environment in a constant temperature test chamber, as shown in Figure 7, with constant humidity and temperature, external dimensions of 10 m × 5 m × 2.8 m, a temperature range of −60 • C to −85 • C, and a water temperature control range of 0.5 • C to −5 • C. The PID touch controller used can not only operate a self-touch digital display but can also write a program to control the test cycle, to achieve high precision temperature control. The spray device, as shown in Figure 8, uses an adjustable nozzle with a spray flow range of 0-60 mm/h.  The heating wire used in the test was a 48 K carbon fiber heating wire, as shown in Figure 9. The heating principle is based on using the internal resistance of the carbon fiber heating wire to generate heat. The carbon fiber heating line has four layers: the first layer is the carbon fiber heating body, the second layer is the electrical insulation material PTFE, the third layer is the high-voltage, corrosion-resistant material polyethylene, and the fourth layer is the PVC layer. The performance parameters of the 48 K carbon fiber heating line are shown in Table 2.  The 48 K carbon fiber hot wire used in this experiment is a new type of hair material, with high mechanical strength and service life, light weight, good softness, good folding resistance, and high resistance to decreases in electric heating efficiency. The performance of 48 K carbon fiber heating wire is shown in Table 2. In addition, it has only a small impact on the concrete body's structural force. Nylon ties have good acid resistance, good corrosion resistance, insulating properties, good age resistance, and strong bearing characteristics, and thus the carbon fiber hot wire U-shaped arrangement, fixed in the upper part of the glass fiber grille, uses nylon ties to fix it, with a spacing of 10 cm.

Temperature Measuring Instruments and Temperature Measuring Lines
The Keysight pyrometer was selected as the temperature measuring instrument. The working temperature range of the detection system is −40 • C to 125 • C. The K-type thermocouple line was selected as the temperature measurement line, and its working temperature is −40 • C-60 • C; the temperature measurement probe is connected to the Keysight thermometer and then connected to the PC side for connection, and the temperature change is recorded in real time. The temperature measuring instrument and the temperature measuring line are shown in Figures 10 and 11, respectively.

Crossbeam Icing Test Conditions
In order to study the effect of ambient temperature and wind speed on the distribution of ice on the surface of the beam and the transfer of heat energy generated by the carbon fiber heating line inside the specimen, the ambient temperature of the test chamber was selected to be −5 • C, −10 • C, and −15 • C; the wind speed was 0 m/s and 4.5 m/s; the rainfall was 30 mm/h; and the spraying time was 1 h for the indoor model tests. The crossbeam icing test working conditions are shown in Table 3.

Crossbeam Ice Melt Test Conditions
In order to study the ice melting pattern of bridge tower crossbeams with built-in carbon fiber heating lines, the ambient temperature of the test chamber was selected to be −5 • C, −10 • C, and −15 • C; the wind speed was 0 m/s and 4.5 m/s; the rainfall was 30 mm/h; and the spraying time was 1 h for the indoor model tests. The beam de-icing test working conditions were designed as shown in Table 4.

Crossbeam Icing Test
In order to investigate the effects of ambient temperature and wind speed on the icing of bridge tower crossbeams, the ambient temperature of the test chamber was selected to be −5 • C, −10 • C, and −15 • C, and the wind speed was 0 m/s and 4.5 m/s for the indoor model tests. The icing characteristics of the crossbeams were obtained through the crossbeam icing tests, and the specific icing locations of the crossbeams were obtained to provide a test basis for subsequent crossbeam ice melting tests.

Influence of Ambient Temperature on Crossbeam Icing
By carrying out model tests for working condition 1, working condition 2, and working condition 3, the icing process of the crossbeam was continuously filmed, and the corresponding ice column histograms were drawn by Origin software, as shown in Figure 12. The first is for the ice on the top surface of the beam, which is sparsely distributed and soft; the second is for the ice layer adhering to the wall surface of the beam, which is hard and dense, but not uniform in thickness; and the third is for the ice column attached to the bottom edge of the beam, which is of varying length and uneven in distribution.
The raindrops form longer icicles at the junction of the sides and bottom of the crossbeam, and it is evident that the icing coverage of the sides of the crossbeam varies at different ambient temperatures, as shown in Figure 13. As can be seen from the figure, at an ambient temperature of −5 • C, the sides of the crossbeam have fewer ice drops, accounting for about 5% of the side area, and the icicles at the junction of the sides and the bottom are shorter and thinner; at an ambient temperature of −10 • C, the sides of the crossbeam have many ice drops, accounting for about 40% of the side area; at an ambient temperature of −15 • C, the sides of the crossbeam are almost covered with ice drops, accounting for about 80% of the side area, and the icicles at the intersection of the sides and the bottom are long and thick. This means that the lower the ambient temperature, the greater the ice coverage on the sides of the beam, making it more difficult to melt the ice.   The maximum, minimum, and average lengths of the ice column at different ambient temperatures for the crossbeam are shown in Table 5, and the dotted line graphs drawn by Origin software are shown in Figure 14. Comparing the test results of condition 1 and condition 2, the maximum, minimum, and average lengths of the ice column increased by 7 cm, 3 cm, and 4.95 cm, respectively, corresponding to growth rates of 53.85%, 60%, and 57.36%, respectively. When comparing the results of the test in condition 2 and condition 3, the maximum, minimum, and average lengths of the ice column increased by 5 cm, 4 cm, and 4.98 cm, respectively, corresponding to growth rates of 25%, 33.33%, and 36.67%, respectively. The results show that as the ambient temperature continued to decrease, the length of the ice column continued to grow, but the growth rate decreased. This is because when the ambient temperature is too low, the rate of ice formation is further accelerated and fewer raindrops flow to the end of the ice column; most of these raindrops freeze midway, so the ice column grows more in the radial direction.

Effect of Wind Speed on Crossbeam Icing
The physical diagram of the crossbeam icing in condition 4 (ambient temperature −15 • C, wind speed 4.5 m/s) is shown in Figure 15, from which it can be seen that the icicles on the crossbeam have a certain inclination due to the wind speed, and the length and thickness of the icicles increase compared with condition 3 (ambient temperature −15 • C, wind speed 0 m/s), while the side ice coverage of the crossbeam reaches 80% and the side ice thickness of the crossbeam increases by 2 cm compared with condition 3. The test results show that the length and thickness of the crossbeam ice column increases with increasing wind speed and that the thickness of the crossbeam side ice layer becomes thicker with increasing wind speed, thus making it more difficult to de-ice.   As can be seen in Figure 16, the length of the ice columns on the beam varies and the distribution of the long and short ice columns is irregular. When the wind speed is applied, the lateral ice coverage of the beam is 80%, but the thickness of the lateral ice on the beam becomes thicker. In addition, the ice column tilts due to the wind, with a tilt angle of approximately 30 • at a wind speed of 4.5 m/s. A comparison of the test results for condition 3 and condition 4 is shown in Table 6, where the maximum, minimum, and average lengths of the crossbeam ice column increase by 9 cm, 8 cm, and 7.78 cm, respectively. The corresponding growth rates are 36%, 66.67%, and 41.92%, respectively. The results show that the application of wind speed results in longer and larger diameter columns of ice on the crossbeam, thicker ice on the sides of the crossbeam, and a certain inclination of the columns, which means that the effect of wind speed is greater than the effect of ambient temperature and more attention should be paid to the effect of wind speed changes on the de-icing of the crossbeam and the prevention of icing.  The crossbeam icing tests revealed that different ambient temperatures and wind speeds have an effect on the distribution of ice columns on the sides of the crossbeam. The thickness of ice on the side of the beam was linearly related to the ambient temperature and wind speed; the higher the wind speed, the greater the thickness of ice on the side of the beam, and similarly, the lower the temperature, the greater the thickness of ice on the side of the beam. The ice thickness on the side of the crossbeam increases with both increasing wind speed and decreasing temperature. The comparison shows that the effect of wind speed on the crossbeam flank ice column is more obvious in the icing test and more attention should be paid to the effect of wind speed changes on crossbeam ice melting. Figure 17 shows the average temperature variation curve of the internal beam model over time, where the dashed horizontal and vertical coordinates represent the time and temperature corresponding to the start of melting ice on the beam, respectively, where the test conditions were: −5 • C, −10 • C, and −15 • C ambient temperature, 0 m/s wind speed, 30 mm/h rainfall, 1 h spraying time, 8 h carbon fiber hot wire energization time, and 8 h temperature measurement by temperature measuring instrument. The HD camera video shows that the ice column on the crossbeam started to melt at 22 min, 75 min, and 141 min during the tests in working condition 5, working condition 6, and working condition 7, respectively. The average temperature of the 12 temperature measurement points in the crossbeam was taken as the average temperature, and the temperatures at which the crossbeam ice started to melt were 1.07 • C, 1.17 • C, and 1.2 • C, respectively.

Influence of Ambient Temperature on Crossbeam Ice Melt
As can be seen from Figure 17, the internal temperature of the beam increased at a faster rate in the time range of 0-4 h in the tests for condition 5, condition 6, and condition 7, with the rate of temperature increase decreasing from that occurring at 0-4 h in the tests for condition 5 and condition 6 at 4-6 h and stabilizing at 6-8 h; the rate of temperature increase began to stabilize in the tests for condition 7 at 4-8 h. The temperature rise test results of the average internal temperature of the electrically heated beam specimens at different ambient temperatures are shown in Table 7, after 8 h of heating, the average temperatures of the temperature measurement points in the tests for condition 5, condition 6, and condition 7 were 20.28 • C, 13.36 • C, and 6.04 • C, respectively. The test data show that the lower the ambient temperature, the lower the temperature reached after the beam was heated and stabilized, and the slower the rate of temperature rise.
This is because the carbon fiber hot wire buried inside the beam model transfers heat to the surrounding area, and group B is midway between groups A and C. Therefore, the temperature of the temperature measurement point of group B is higher than that of both groups A and C. As group C is on the lower surface of the side of the beam, where the root of the icicle is mainly formed, group C best represents the lowest temperature of the surface of the carbon fiber hot wire under ideal conditions.

Effect of Wind Speed on Crossbeam Ice Melt
The effect of wind speed on the de-icing of the crossbeam is analyzed by comparing the test results of working condition 8, working condition 9, and working condition 10 with those of working condition 5, working condition 6, and working condition 7. Figure 18 shows the variation curves of the average internal temperature of the beam model with time at ambient temperatures of −5 • C, −10 • C, and −15 • C, where the dashed horizontal and vertical coordinates represent the time and temperature corresponding to the start of melting ice on the beam, respectively, where the test conditions were as follows: wind speed of 4.5 m/s, rainfall of 30 mm/h, carbon fiber heating line energized for 1 h, and temperature measurement for 8 h. The temperature rise test results for the average internal temperature of the electrically heated beam model at different ambient temperatures are shown in Table 8. The HD camera video shows that the ice column on the crossbeam started to melt at 22 min, 75 min, and 141 min in the tests for condition 8, condition 9, and condition 10, respectively, and the average critical temperatures at which the crossbeam started to melt were 1.28 • C, 1.31 • C, and 1.34 • C, respectively.
The graphs of the changes in the internal temperature field of the beams in test conditions 8, 9, and 10 are shown in Figure 18. It can be seen from the graph that in the three sets of different ambient temperature tests, the temperature rise rate of the internal temperature of the beam was faster in the time range of 0-2 h. In the eighth set of tests in the time range of 2-8 h, the temperature rise rate decreased compared with that occurring at 0-2 h; in the ninth set of tests in the time range of 2-6 h, the temperature rise rate decreased compared with that occurring at 0-2 h; and in the time range of 6-8 h, the temperature rise rate tended to be stable. In the tenth set of tests in the time range of 2-3 h, the temperature rise rate decreased compared with that occurring at 0-2 h, and in the time range of 3-8 h, the temperature rise rate tended to be stable. In the tenth group of trials, the rate of temperature rise decreased from that occurring at 0-2 h in the time period at 2-3 h and levelled off at 3-8 h.
The corresponding mean temperatures at 2 h of heating were 5.84 • C, 3.07 • C, and −0.97 • C, respectively. The ambient temperature of −15 • C in the test group started to stabilize after 2 h of heating, whereas the other two test groups were still in slow warming at this time. The average critical temperatures for ice melt in working condition 5, working condition 6, and working condition 7 were 1.07 • C, 1.17 • C, and 1.2 • C, respectively, whereas the average critical temperatures for ice melt in working condition 8, working condition 9, and working condition 10 were 1.28 • C, 1.31 • C, and 1.34 • C, respectively. After 8 h of heating, the average temperatures of the beams in working condition 8, working condition 9, and working condition 10 were 10.38 • C, 7.94 • C, and 1.37 • C, respectively. Compared with the test results of working condition 5, working condition 6, and working condition 7, the temperature values after heating and stabilization were reduced by 9.9 • C, 5.42 • C, and 4.67 • C, respectively. This is due to the wind speed increasing the heat loss and reducing the heat transfer efficiency of the carbon fiber heating line. The results show that the critical temperature of the crossbeam ice melt is also affected by the wind speed: the higher the wind speed, the higher the critical temperature of the crossbeam ice melt. In addition, the temperature at the beginning of the crossbeam ice melt shows a positive correlation with the wind speed.    Figure 18 shows that at an ambient temperature of −15 • C and a wind speed of 4.5 m/s, the average internal temperature of the beam after heating and stabilization is 1.37 • C, which is higher than the average critical temperature of 1.31 • C for the beam to melt ice under this condition. This is sufficient to melt the ice, which proves the feasibility of this solution and can provide a reliable reference for the practical engineering application of anti-icing mechanisms in bridge tower crossbeams.

Conclusions
In this study, we enhanced the temperature of a bridge beam structure using an electric heating method to achieve the effect of melting snow and melting ice. Considering factors such as the ambient temperature and wind speed, we conducted experimental research on the icing and de-icing of a concrete bridge tower bridge beam model with pre-buried carbon fiber heating line. The main conclusions are as follows: (1) This paper proposes a method for melting snow and ice on bridge tower crossbeams with pre-buried carbon fiber hot wire, which is fast, accurate, stable, energy-saving, and efficient. (2) The ambient temperature and wind speed will affect beam icing. Decreasing the ambient temperature will increase the length of the icicles, whereas increasing the wind speed will increase the length and thickness of the icicles, thus exacerbating the potential threat. (3) The lower the ambient temperature, the higher the ice coverage on the side of the beam, and the higher the adhesion force, making it more difficult to melt the ice. The critical temperature of the bridge beam ice melt is influenced by the ambient temperature and wind speed; this is negatively correlated with the ambient temperature and positively correlated with the wind speed. (4) At an ambient temperature of −15 • C and a wind speed of 4.5 m/s, the average internal temperature of the beam after heating and stabilization is 1.37 • C, which is higher than the average critical temperature of 1.31 • C for the beam to melt ice under this condition. This temperature is sufficient to melt ice, verifying the feasibility of this solution and is a reliable reference for the practical application of ice melting in bridge tower beams.
This study has certain limitations, because this test was carried out in the laboratory, and only the factors of ambient temperature and wind speed were considered. In actual bridge engineering, there are many other environmental factors that have not been considered. For example, the influence of environmental humidity; different environmental humidity will affect the icing and deicing of bridge beams. Therefore, the results of this study only provided a theoretical basis for reference.
In future work, we plan to add icing prevention tests to the bridge beam, i.e., operating carbon fiber heating wires before low-temperature rainfall, so that the surface temperature of the beam cannot be frozen. In addition, we plan to conduct numerical simulation analysis to de-icing tests on a bridge beam, as well as an ice prevention test. We also plan to add more environmental factors to enrich the content. For example, by increasing the power rating of the carbon fiber hair wire, the de-icing efficiency of the bridge beam can be accelerated to a greater extent.