Wind Tunnel Test of Icing Distribution on the Leading Edge of a 2D Blade Airfoil for Wind Turbines

Abstract

As clean and renewable energy, wind energy has been widely used in the world. The wind turbine is a kind of rotating machinery, which can convert wind energy into mechanical energy and electrical energy. Wind turbines sometimes face a variety of extreme weather conditions, such as icing, heavy snow, lightning, sand storm, and so on, which affect the safety operation of wind turbines. In the present study, icing events on the blade surfaces of wind turbines are focused, and the wind tunnel test of icing was carried out on a 2D blade with NACA0018 airfoil used in wind turbines. In the icing tests, three kinds of ambient temperatures were selected, including −5 °C, −10 °C, and −15 °C, and two kinds of wind speeds were decided, including 5m/s and 10m/s. The icing distributions on the blade surface at the different attack angles were tested and recorded by a high-speed camera for several minutes. The ice accretion and distribution characteristics at the leading edge of the 2D blade airfoil were acquired and analyzed. The findings can provide a reference for the research on the icing mechanism and the de- and anti-icing of wind turbines.

1. Introduction

As clean and renewable energy, wind energy has been widely used in the world and will be paid more and more attention [1]. A wind turbine is the most important equipment for the utilization of wind energy, and it is also a kind of rotating machinery, which can convert wind energy into mechanical energy, heat energy, and electrical energy [2]. According to the rotation mode of a wind turbine shaft, there are two key kinds of wind turbines, which are the horizontal axis wind turbine (HAWT) and the vertical axis wind turbine (VAWT). Currently, the HAWTs are used for the large-scale wind turbines mostly. In contrast, the VAWTs are often used for the small-scale wind energy utilization [3]. Thanks to years of researching, the performance of wind turbines has been greatly improved. The research focus also extends from the improvement of efficiency to how to ensure the safe, stable, and efficient operation of wind turbines. Meanwhile, global climate change intensifies and extreme weather occurs more frequently. Wind turbines are facing more severe weather conditions, such as typhoon, earthquake, lightning, blizzard, icing, etc. [4]. In the cold and moist regions, wind turbines will face the icing problem. When the environment meets a certain condition, supercooled water droplets in the air will accrete on the wind turbine surface and become icing [5]. An icing photo of a wind turbine taken by us in Northeast China in December 2019 is shown in Figure 1.

Figure 1.

Icing on the blades of HAWT.

Icing event occurring on the wind turbine blade surface will greatly degrade the efficiency and safety of operation. The research contents of icing mainly include distribution and characteristics of icing on a blade [6, 7, 8], aerodynamic performance effect of icing on blade and rotor [9, 10, 11], anti-icing and de-icing methods and technologies [12, 13, 14, 15], icing detection technologies [16], etc. There are two main icing research methods, numerical simulation [17, 18] and icing wind tunnel test [19, 20]. In the present study, a small-scale icing wind tunnel was used to obtain the distribution characteristics of icing on a model of 2D blade with NACA0018 airfoil. Under different ambient temperatures and wind speeds, experiments on ice accretion on the 2D blade at two kinds of angel of attacks for 4 minutes were carried out. The icing distributions were recorded and analyzed. This research can provide a reference for researching icing mechanism and de- and anti-icing technologies of wind turbines.

2. Methods

2.1 Experimental system

Figure 2 shows an icing wind tunnel experimental system in Northeast Agricultural University used in this study. It’s a low-speed and return-flow type wind tunnel. The refrigeration and spray devices were installed in a normal wind tunnel to provide low temperature and icing environment. The detailed information about this system can be found in the reference [21]. The test section is 250 mm × 250 mm. The range of wind speed is from 1 m/s to 20 m/s. The ambient temperature can be controlled lower to −20°C. By using different sprayers, the icing wind tunnel can supply different kinds of liquid water content (LWC) and medium volume diameter (MVD) of the supercooled water droplet. The LWC is from 0.1 g/m³ ~ 5 g/m³, and the MVD is from 20 μm ~ 100 μm.

Figure 2.

Icing wind tunnel experimental system used in this study. (a) Schematic diagram of the icing wind tunnel experimental system. (b) Photo of the icing wind tunnel.

2.2 Test model

Figure 3 shows the test model of blade segment with airfoil profile used in wind turbine. To obtain the basic and normal icing characteristics of wind turbine blade, the symmetrical airfoil of NACA0018 was selected because it is often used for wind turbines and basic researches. The blade material was the aluminum alloy, which is usually selected for icing wind tunnel tests because of its good and stable thermal conductivity. According to the size of test section of wind tunnel, the test model has the chord length of 150 mm and the thickness or spanwise length of 20 mm. For this test model, the airfoil profile is consistent along spanwise, and its thickness is small. Therefore, it can be seen as a 2D blade airfoil in the present experimental research. The 3D effect of it can be negligible.

Figure 3.

Test model of the blade with NACA0018 airfoil.

2.3 Test conditions

Based on the research object, several test parameters were selected for comparison. It included ambient temperatures, wind speeds, and icing times. To check the effect of an angle of attack (AOA) on the blade icing distribution, two kinds of angle of attacks (AOAs), such as 0°and 10°, were selected. The test conditions are listed in Table 1. Furthermore, the LWC in the icing wind tunnel used in the present study was in the range of 0.8 ~ 1.6 g/m³. The MVD was about 65 μm. In this study, the effects of LWC and MVD on icing of blade airfoil were not researched, which will be carried out deeply in the follow-up study. When the spray system begins to work, a high-speed camera (Phantom v5.1, with the revolution ratio of 1024 × 1024 pixels) begins to record the process of icing on blade surface and take photos at the moments of 2 minutes and the 4 minutes. By using the drawing software, the profile of icing shape can be obtained and analyzed.

Angel of attack (°)	Wind speed (m/s)	Temperature (°C)	Icing time (min)
0	5, 10	−5, −10, −15	0, 2, 4
10	5, 10	−5, −10, −15	0, 2, 4

Table 1.

Experimental conditions of the icing tests in the icing wind tunnel.

3. Results and discussion

3.1 Icing distributions on the leading edge of blade surface

Figures 4 and 5 show the photos of icing distribution at the leading edge of blade under all test conditions. According to the figures, it can be seen that icing occurs on the leading edge of blade for all test conditions. It means that icing will appear when certain environmental conditions are met. However, the characteristics of icing distribution and ice accretion are obviously different under different test conditions.

Figure 4.

Photos of icing distribution on blade at the wind speed of 5 m/s.

Figure 5.

Photos of icing distribution on blade at the wind speed of 10 m/s.

For the effects of ambient temperature, the most typical characteristic is the change of icing type. When the temperature was −5°C, the type of icing on the blade surface was glaze ice with a transparent color. With the decrease in temperature, the type of rime ice, whose color was white, gradually began to appear. At the temperature of −10°C, there was a little rime ice, which only appeared on the upper surface away from the leading edge. This phenomenon was more obvious at the AOA of 0° than the one at 10°. When the temperature reached to −15°C, most of the ice transformed into the rime ice, while there was very little glaze ice, which only appeared at the front of the leading edge. The glaze ice was just formed by condensation of newly flowing supercooled water droplets in the ultimate period of photography. These results indicate that the icing type on the leading edge of blade is glaze at high temperatures, such as above −5°C, and transforms into mixed ice and rime ice with the decrease in temperature. Furthermore, another phenomenon should be mentioned. For the AOA of blade at 0°, when the temperature was −5°C, there was a short icicle appearing on the lower surface near the leading edge of the blade at the icing time of 4 minutes. The reason for an explanation of this result is that the heat transfer rate of water droplets decreased with the increase in icing time. At the initial icing stage, the water droplets impinged directly on the blade surface whose temperature was the same as the environmental temperature. In this case, the water droplets froze in a short time because of higher heat conductivity coefficient of aluminum material (237 W/m·K) and low blade surface temperature. That is why the type of rime ice is generated on the blade surface. After that, there was a layer of ice covering the blade surface. The consequent water droplets impinged on the ice layer. The water droplets could not freeze in a short time because of low heat conductivity of ice (2.22 W/m·K). Some super-cooled water droplets attaching on the ice layer surface had not enough time to freeze into ice completely due to the high temperature. These water droplets ran back along the blade surface to the trailing edge, and the icicle was generated. In this case, the type of icing was glaze ice due to high temperature of the icing blade. Under the action of wind speed and gravity, these incompletely frozen water droplets grow and form an ice icicle obliquely downward.

For the effects of wind speed, it was found that the wind speed does not play a key role in the change of icing type. When the temperature was constant, the type of icing was basically the same. When the wind speed increased from 5 m/s to 10 m/s, the icing amount decreases slightly. Meanwhile, the icing shape became smoother, especially at the temperature of −15°C. This may be caused by the constant flow rate of the spray nozzle in unit time. In the present study, the pressure of the pump is constant, which makes the flow rate of the nozzle constant in spite of increasing wind speed. Therefore, the water content passing through a cross-section is constant in unit time. In theory, the icing amounts are the same under two kinds of wind speeds. However, when the wind speed is higher, more water droplets with lightweight are blown away by high-speed wind, and the amount of water droplets impinging on the blade decreases. Additionally, the heat exchange of water droplets with air also accelerates and enhances because of the increase in wind speed. In this case, some small water droplets freeze before impinging on the blade surface, which also results in the decrease in the icing amount.

For the different AOAs, the amount of icing on the lower surface of blade increased because there was an upward angle of attack against wind. In contrast, the amount of icing on the upper surface had a decreasing trend. The key reason is that the windward state of the blade changes from symmetry to asymmetry due to the existence of AOA, which leads to the asymmetry of icing distribution on the leading edge of blade. Therefore, it was concluded that the AOA, or the windward state of blade, determines the location and distribution of icing on blade surface. Furthermore, a deep discussion can be made on how to use these results to the real wind turbine blade. As is known to all, the wind turbine blades always operate in a rotating state. Although the blade model in this research was in a static state, the research findings can be used for the analysis of rotating blades. Based upon the aerodynamics of wind turbine blade, the wind speed and the rotational speed will form a resultant speed at the local element of blade airfoil, which is known as relative wind speed. The AOA in the rotating condition should be decided by the relative wind speed. Therefore, the research findings obtained in the static condition can be used for the analysis of rotating blades.

3.2 Icing profile and icing area

In order to quantitatively analyze the distribution characteristics of icing, the captured icing photos were processed and the icing contours at the different moments were obtained, which are shown in Figures 6 and 7.

Figure 6.

Ice accretion on blade at the wind speed of 5 m/s.

Figure 7.

Ice accretion on blade at the wind speed of 10 m/s.

For comparatively analyzing with the results at the AOA of 0 degrees, the figures of the icing blade at the AOA of 10° are turned back 10 degrees and located at horizontal level for convenient observation and comparison.

Based on these pictures showing the profile of icing, the change of icing shape can be seen clearly. Furthermore, from these figures, the icing area (A_i) is calculated in the present study. The schematic diagram of icing area is shown in Figure 8.

Figure 8.

Schematic diagram of icing area.

Figure 9 shows the changes in the icing area with the increase in icing time under all test conditions. Additionally, a dimensionless parameter, icing area ratio (η_s), can also be defined to quantitatively analyze the icing area. It is defined by the ratio of icing area (A_i) and the blade airfoil area (A), which is expressed in Eq. (1).

Figure 9.

Variation of icing area with icing time.

$${\eta }_s=\frac{A_i}{A}\times 100\%$$E1

The variations of icing area ratios with icing time, calculated from the data in Figure 9, are shown in Figure 10.

Figure 10.

Variation of icing area ratio with icing time.

As shown in Figure 9, the icing area increased linearly with the increase in icing time in general. This characteristic can also be found in the result of the icing area ratio in Figure 10. Therefore, in this section, it only focuses on the analysis of icing area ratio in Figure 10. In the period of 4 minutes icing time, for all conditions in this research, the maximum icing area ratios reached to 6.79% and 8.31% at the AOAs of 0° and 10°, respectively. At the wind speed of 5 m/s, with the decrease in temperature, the growth rate of icing area ratio increased. This phenomenon was obvious, especially at the AOA of 10°. However, for the wind speed of 10 m/s, the effect of temperature on growth rate of icing area ratio became more obvious than that at a wind speed of 5 m/s. The reasons for this result are multifaceted and complex. The high wind speed gives high kinetic energy to the supercooled water droplets, which makes the impact course of water droplets on the blade surface complicated. Also, higher wind speed will affect the characteristics of heat transfer more obviously in comparison with that of the lower wind speed. When the temperature is low, a higher wind speed intensifies the heat transfer of water droplets, which enhances the growth rate of icing area. Deep researches on this issue will be carried out in our further tests.

More discussions were given on the effects of icing time and angle of attack. For the icing time, the icing time length of 4 minutes is selected in this study. Of course, with the increase in icing time, the icing area will increase. In nature, the process of icing on a wind turbine is a long course, which takes time from hours to days. In this research, the main purpose is to explore the initial stage of icing on the leading edge of blade surface. When the ice forms to a certain shape and thickness, the consequent icing process occurs between supercooled water droplets and ice formed on the blade surface, not the blade surface. For the angle of attack, the most important effect is to change the windward area of the blade. Based on the test results of AOA, it is concluded that the icing area will increase along with AOA. For the blade in a static condition, the range of AOA is wide. However, for the rotating blade of a real wind turbine, the relative angle of attack is controlled in a limited range. Therefore, the test results can be a useful reference to the icing research of a real wind turbine.

4. Conclusions

In this research, the icing tests of a blade model for a wind turbine were carried out in an icing wind tunnel. The distribution characteristics of icing on the leading edge of the blade were obtained and analyzed. Some key results in the present study are summarized and listed as follows:

1. In the present study, the icing on the leading edge of a blade with NACA0018 airfoil was explored and obtained. It is concluded that the ice type has good relationship with the ambient temperature. With the decrease in temperature, the ice type changes from glaze ice, mixed ice, to rime ice. Meanwhile, the phenomenon is more obvious when the wind speed increases. Additionally, with the increase in wind speed, the icing amount of the blade decreases slightly as the flow rate of spray nozzle is constant. Under low-temperature condition, high wind speed can accelerate the growth rate of icing.

2. When the environmental condition keeps constant, the icing area increases linearly along with the icing time. Additionally, the icing area is also affected by the angle of attack. With the increase in angle of attack, the windward area of the blade increases, which results in an increase in the icing area. Moreover, the area covered by ice on the lower surface increases with respect to the AOA. The maximum icing area ratios reach to 6.79% and 8.31% at the AOAs of 5° and 10°, respectively.

Based on the method of icing wind tunnel test in this research, the characteristics of icing on blade surface can be researched and obtained. This study can be a reference to the icing mechanism research and the anti- and de-icing research of wind turbines.

Acknowledgments

This work is supported by the National Natural Science Foundation of China (NSFC) [grant number 51976029]. The authors would like to thanks for the support.