2025 Volume 66 Issue 3 Pages 201-207
Reducing aerodynamic noise emitted by pantographs is one of the key challenges to be overcome to reduce the environmental impact of railways and enable faster train running speeds. Previous studies proposed a method for reducing aerodynamic noise by applying porous material to pantograph head support covers. This study proposes a new practical method to achieve the same aerodynamic noise reduction effect as using porous material. Wind tunnel test results showed that this new method can reduce aerodynamic noise to almost the same extent as using porous material.
To achieve higher speed on Shinkansen, it is important to reduce noise generated by trains, particularly aerodynamic noise which increases proportionally from the sixth to the eighth power of the train speed. Among Shinkansen vehicle components responsible for noise, pantographs are one of the main sources of aerodynamic noise. Pantograph heads and head supports (Fig. 1) in particular are known to be major contributors to this noise.
Pantograph heads, which slide against the overhead wire, in particular, are a challenge. This is because not only is it necessary to reduce aerodynamic noise, it is also essential to stabilize lift characteristics (i.e., lift characteristics must not change significantly with variations in the direction of the oncoming wind or changes in contact strip shape caused by wear). Consequently, the most common cross-sectional shape of pantographs is rectangular as shown in Fig. 2(a) (henceforth referred to as “current pantograph head”). Given these considerations, previous research used a combination of CFD (Computational Fluid Dynamics) analysis and shape optimization to propose a pantograph head cross-sectional shape that reduces aerodynamic noise and stabilizes lift characteristics based on the assumption of unidirectional running [1], and also developed a multi-segmented smooth-profile pantograph head with a suspension mechanism based on that cross-sectional shape [2, 3] (Fig. 2(b), Fig. 3).
Another important aspect in reducing aerodynamic noise generated by pantograph heads and head supports is controlling the flow interference caused by these parts. Based on the assumption of running in the knuckle forward direction, a previous study proposed aerodynamic noise reduction measures, such as an improved head support that reduces flow interference by offsetting the pantograph head's installation position upstream and applying a porous material to the cover of the head support [3] (Fig. 3). Of these aerodynamic noise reduction measures, the application of a porous material is straightforward as it only changes the surface properties of the member without significantly changing its shape or structure. The structural characteristics of the porous material also necessitate adhesive installation, but research has confirmed that the installation strength was not inhibited in cases of application to actual vehicles [4], and this is considered to be a technology that has a certain prospect for practical application. However, the porous material may be crushed or deformed when struck by a flying object, and replacing only the porous part is difficult because it is attached with an adhesive. Therefore, to ensure the practical application of aerodynamic noise reduction methods by improving the surface properties of members, such as porous materials, it is essential to improve the strength and maintainability of these members. In some cases, new methods that do not rely on porous materials may be needed. To that end, in this study, we propose a flow bypass method as a practical alternative method that can reduce aerodynamic noise based on the same principle as that of porous materials and improve the strength and maintainability of members. We confirmed the aerodynamic noise reduction effect in a wind tunnel test.
A porous material has countless pores, as shown in Fig. 4, and a structure in which adjacent pores are connected to each other is known to be necessary for reducing aerodynamic noise. Previous research has also clarified that the aerodynamic noise reduction effect of porous materials is not due to acoustic effects such as sound absorption, but rather due to a fluid dynamic effect in which the flow field is stabilized by the natural inflow and outflow of air that occur on the surface of the porous material [5]. Therefore, designing a structure in which natural inflow and outflow of air occurs according to the pressure differential on the member surface is important for achieving the effect of porous materials by another method.
In this study, we assume that a metal porous material is applied to a head support part as shown in Fig. 3, and propose an alternative method to the porous material, in which the flow is bypassed from the upper surface to the side by providing a flow channel at the corner of the upper side of the head support (flow bypass method). Figure 5(a) shows an overview of this method. This method involves connecting the upper side of the head support's upper corner, where the flow stagnates and the pressure rises, to the lateral side, where the flow separates and the pressure falls, with a bypass channel. The aim is to generate a natural inflow and outflow from the upper side to the lateral side.
The test sample was a currently used head support cover widened by 10 mm in the lateral direction and a removable member with a bypass channel at the corner. The area to which the bypass channel is applied generally has a width W = 10 mm, height H = 10 mm, and length L = 128.5 mm in the ridge line direction. As shown in Fig. 5(b), the bypass channel is generally provided in the two directions of “direction 1” perpendicular to the upper surface of the apex and “direction 2” perpendicular to the lateral side, with some tests adding a bypass channel in “direction 3” parallel to the ridge.
The wind tunnel test was conducted in the Railway Technical Research Institute's large-scale low-noise wind tunnel (open-type measurement section, air outlet 3 m × 2.5 m, maximum wind speed 111 m/s) using a real pantograph for Shinkansen trains as the test sample. Figure 6 shows the wind tunnel test conditions. The pantograph head was a multi-segmented smooth-profile pantograph head [2, 3], shown in Figs. 2(b) and 3, and the pantograph was installed in the knuckle forward direction. The test wind speed was 100 m/s (360 km/h), and the Reynolds number during the test was 6.5 × 105, which was calculated using the lateral width of the head support as the representative length. The main purpose of this test was to evaluate the aerodynamic noise of the head support and to minimize the noise emitted from other parts, the test was generally performed with all tap holes and movable gaps in the multi-segmented smooth-profile pantograph head sealed off with tape (Fig. 6(b)).
The aerodynamic noise measurement was conducted under conditions where the entire pantograph was exposed to the airflow, and the aerodynamic noise was evaluated using an omnidirectional microphone installed diagonally below the side of the pantograph. The omnidirectional microphone was installed 356 mm upstream from the center of the pantograph head in the flow direction, 5 m to the side of the center of the pantograph head in the lateral direction, and 2.6 m below the upper surface of the pantograph head in the vertical direction. The noise level was calculated by conducting 1/3 octave band frequency analysis on the aerodynamic noise measurement waveform and applying A-weighting correction. The overall value was generally evaluated using the partial overall value (“POA value”) calculated from the frequency band of 400 Hz or more, which is the frequency band where the contribution of the head support is dominant. Furthermore, conditions equivalent to background noise were set by conducting tests with the pantograph head removed in order to prevent the influence of flow interference with the pantograph head and head support, and these results were shown with the aerodynamic noise measurement results as needed.
First, the test conditions for the four types of head support shown in Fig. 7 were set to confirm the effect of applying bypass channels and the influence of the internal bypass channel configuration [6]. In this study, the cover of the currently used head support (Fig. 7(a) hereinafter “currently used”) was widened by 10 mm in both the left and right directions (Fig. 7(b) hereinafter “widened”), and a bypass channel was introduced in the corner of the “widened” head cover (Fig. 7(c), hereinafter “bypass”). Furthermore, as a comparison of the aerodynamic noise reduction effect, aerodynamic noise measurements were conducted for the condition where a urethane porous material (Inoac Molto Filter MF-13) was applied to the corner of the “widened” head cover (Fig. 7(d), hereinafter “porous”). Here, for “bypass,” circular bypass channels with a diameter of 2 mm were arranged with a center-to-center distance of 4 mm, and three conditions for the direction of the bypass channels were set as shown in Fig. 8. Specifically, the following three conditions were set: “2-axes” (Fig. 8(a)), in which the bypass channels were made in direction 1 and direction 2; “3-axes” (Fig. 8(b)), in which a bypass channel in direction 3 was added to “2-axes” (Fig. 8(b)); and “cavity” (Fig. 8(c)), in which a bypass channel in direction 3 had a rectangular cross-section of 6 mm × 6 mm. For all bypass channel conditions, the members with bypass channels were made using a 3D printer. Furthermore, given the constraints of dividing the parts that constitute the corners, the external dimensions of the porous material were set to W = 10 mm, H = 16.7 mm, and L = 139.1 mm (see Fig. 5(a) for the parts marked with each symbol), which are slightly larger than the applicable area of the bypass channels.
Figure 9 shows the aerodynamic noise measurement results of this study. Referencing the comparison diagram of the POA values (Fig. 9(a)), the POA values of the “currently used” and “widened” support covers were almost identical, and we can confirm that the impact of widening the support cover on the aerodynamic noise was small. Given this, POA values when various bypass channels are made were compared, using “currently used” as the reference, showed an aerodynamic noise reduction effect of 0.6-0.7 dB, where it could be confirmed that the difference in aerodynamic noise due to the internal bypass channel configuration was also small. However, for the “porous” case, the POA value was reduced by 0.9 dB. Therefore, the results showed that the aerodynamic noise reduction effect of making the bypass channel was slightly smaller than that when porous material was applied.
Next, focusing on the comparison diagram of the aerodynamic noise spectrum (Fig. 9(b)), conditions where various bypass channels are applied to the “currently used” system showed that the noise level was reduced over a wide frequency band from 400 Hz to 8 kHz, but the aerodynamic noise reduction effect in each frequency band was slightly smaller than that of the “porous” system. Although not at a level that considerably influences the POA value, when focusing on the 12.5 kHz band, we could observe differences depending on the configuration of the internal bypass channel: “2-axes,” which omits the bypass channel in direction 3, had a slight peak noise, whereas “3-axes” and “cavity”, which add a bypass channel in direction 3, generated no peak noise. The weak peak noise at 12.5 kHz generated in the “2-axes” system was separately confirmed to have no wind speed dependency in its frequency, and because 1/4 of the wavelength of the sound wave (6.8 mm) approximately coincided with the bypass channel depth of 8 mm, this was thought to be air resonance sound with the bypass channel depth being 1/4 of the wavelength. It was also inferred that this peak noise could be reduced by providing a bypass channel in direction 3.
4.2 Effects of bypass channel shape and opening ratioThe results presented in the previous section confirmed that aerodynamic noise can be reduced with a bypass at the corner of the support cover, but the reduction in aerodynamic noise was smaller than that achieved when a porous material was used. This was thought to be due to the insufficient flow rate of the inflow and outflow from the upper surface of the head support to the lateral surface. Therefore, we sought to increase the flow rate of the inflow and outflow by investigating the effect of changes in the cross-sectional shape and size of the bypass channel on aerodynamic noise reduction [7].
Figure 10 shows the opening conditions of the bypass channel. In this section, the internal bypass channel configuration was the same as the “3-axes” system (Fig. 8(b)) described in the previous section, the center-to-center distance of the bypass channels was fixed at 4 mm, and tests were conducted under four conditions: a circular bypass channel with a diameter of 2 mm, as described in the previous section (Fig. 10(a), henceforth “circular (2 mm)”), a square channel with a cross-sectional area of 2 mm on each side (Fig. 10(b), “square (2 mm)”), and two cases where the diameter and side of the bypass channel were respectively expanded to 3 mm (Figs. 10(c), (d), henceforth, “circular (3 mm)” and “square (3 mm)”). For all bypass channel conditions, the members with bypass channels were made using a 3D printer. The conditions for the “currently used” and “porous” used for comparison were the same as those described in the previous section.
Figure 11 shows the results of the aerodynamic noise measurements. Figure 11(a) plots the POA value with respect to the opening ratio of the bypass channel, where the opening ratio is calculated from the total opening area of the bypass channel relative to the bypass application area on the upper surface of the corner (10 mm × 128.5 mm). This figure shows that the POA value tended to decrease with increasing opening ratio of the bypass channel, and that among the samples evaluated in this study, those with an opening ratio of approximately 35% could achieve the same aerodynamic noise reduction effect as that obtained when a porous material is applied.
Referencing the comparison diagram of the aerodynamic noise spectrum (Fig. 11(b)), we could confirm that the noise level was reduced over a wide frequency range from 400 Hz to 8 kHz by applying the bypass channel, as stated in the previous section, and that the aerodynamic noise reduction effect was almost equal to that of the “porous” material for the “circular (3 mm)” and “square (3 mm)” materials with large opening ratios. No significant differences were observed between circular and square channels with similar cross-sectional dimensions, except for slight differences in the POA value and aerodynamic noise spectrum due to differences in the opening ratio. Therefore, the results indicate that the aerodynamic noise reduction effect does not depend significantly on the cross-sectional shape of the bypass channel.
4.3 Aerodynamic noise reduction effect of prototype with actual specificationsGiven the study results up to the previous section, the device shown in Fig. 12 was prototyped as an aerodynamic noise reduction device with actual specifications, where we confirmed its aerodynamic noise reduction effect [8]. This device was constructed by making channels in an aluminum plate with a thickness of 10 mm, considering the strength of the members, and the area of the channel was the same as that described in the previous section, with a width W = 10 mm and height H = 10 mm, but the length in the ridge direction was slightly reduced to L = 121.5 mm owing to workability constraints. The results of Section 4.1 confirmed that the direction of the flow channel did not considerably influence the aerodynamic noise reduction effect, so a 2-axis type was selected in which the flow channels were made only in the two directions of “direction 1” and “direction 2,” considering the workability of the member. As the results of Section 4.2 also indicated that the size and shape of the bypass channels must ensure an opening ratio of approximately 35%, circular bypass channels with a diameter of 3 mm were arranged with a center-to-center distance of 4 mm. This device was designed to be installed by fastening it to the head support. The device can be independently removed for inspection and replacement during repair, so maintenance is easier than that with porous materials that require adhesive installation. The mass is approximately 300 g for both sides, which is sufficiently small compared to the equivalent mass of the upper frame and is not expected to have a large impact on the current collection performance.
We evaluated the aerodynamic noise reduction effect by applying this method to three types of pantograph head and head support conditions, and the aerodynamic noise reduction effect was compared with that of a support cover with metal porous material developed in a previous study [3] (Fig. 3, area covered by porous material was width W = 10 mm, height H = 40 mm, and length L = 140.3 mm in the ridge direction). The pantograph head and head support conditions were as follows: “condition 1,” where both the pantograph head and head support were the currently used products; “condition 2,” where only the pantograph head was changed to a multi-segmented smooth-profile pantograph head in “condition 1”; and “condition 3,” where an improved head support was applied to “condition 2.” For all conditions, the reference condition of the support cover was the currently used support cover. The multi-segmented smooth-profile pantograph head and improved head support were modified from those proposed in previous research (Fig. 3), with slight changes to the external shape [9]. Furthermore, in this section, we sought to evaluate the aerodynamic noise of the pantograph during actual running by sealing the gaps between the pantograph head elements of the multi-segmented smooth-profile pantograph head with soft fur material [3] instead of aluminum tape in order to allow for mobility, and the overall noise was evaluated using the OA value instead of the POA value. This resulted in a slightly smaller aerodynamic noise reduction effect than that indicated in the previous section.
Figure 13 shows the aerodynamic noise measurement results. This figure confirms that the aerodynamic noise was reduced over a wide frequency range above 630 Hz by applying the proposed device (bypass channel) for all the pantograph head and head support conditions. The reduction in the OA value obtained using this device was 0.3-0.4 dB, which was slightly smaller than the aerodynamic noise reduction effect of the porous material (0.4-0.5 dB), but it could be said that almost the same aerodynamic noise reduction effect was obtained. Although the effect on the OA value was small, the results showed that when applying this device a weak peak noise in the 12.5 kHz band was generated. This peak noise was inferred to be the air resonance sound that is specific to the 2-axis type channel configuration confirmed in Section 4.1. In the study conducted in this section, where the opening ratio of the channel was increased in comparison to the study conducted in Section 4.1, the peak noise was thought to have occurred more prominently with increasing flow rate through the bypass channel.
The above results confirmed that the proposed device generated a weak peak noise in the 12.5 kHz band, but that it could achieve an aerodynamic noise reduction effect that was almost equivalent to that of a porous material applied under various pantograph head and head support conditions.
In this study, we proposed a method for bypassing the flow in high-speed train pantographs by inserting a flow channel at the corner of a member. It is a practical alternative method to porous materials that is based on its aerodynamic noise reduction mechanism. This achieved a similar effect to the porous material while improving member strength and maintainability. We also conducted a wind tunnel test to confirm the aerodynamic noise reduction effect. From the results, the following conclusions were drawn:
(1) Aerodynamic noise was reduced by making a bypass channel connecting the upper and lateral surfaces of the head support.
(2) Studies on the direction of the bypass channels showed no large differences in the aerodynamic noise reduction effect between a 2-axes type with a bypass channel perpendicular to the upper and lateral surfaces, and a 3-axis type with an additional bypass channel in the ridge direction. However, the 2-axes type generated a weak peak noise that was inferred to be the air resonance sound related to the depth dimension of the bypass channel.
(3) Studies on the opening ratio of the bypass channel showed that among the samples evaluated in this study, an opening ratio of approximately 35% could achieve an aerodynamic noise reduction effect that was approximately equivalent to that of porous material.
(4) We developed a device with bypass channels in an aluminum plate as an aerodynamic noise reduction device with actual specifications, and we evaluated its aerodynamic noise reduction effect. Results showed that an aerodynamic noise reduction effect of 0.3-0.4 dB was achieved, which was approximately equivalent to that of porous materials, for three representative types of pantograph head and head support conditions.
In addition to pantographs, this method could be applied to a variety of equipment and members as a practical aerodynamic noise reduction measure that could replace porous materials.
|
Takeshi MITSUMOJI Senior Researcher, Current Collection Laboratory, Railway Dynamics Division Research Areas: Aerodynamics of Pantograph |
|
Yuki AMANO, Ph.D. Researcher, Current Collection Laboratory, Railway Dynamics Division Research Areas: Catenary-Pantograph Interaction, Dynamics of Machinery, Mechanical Vibration |
|
Mariko AKUTSU, Ph.D. Assistant Senior Researcher, Noise Analysis Laboratory, Environmental Engineering Division Research Areas: Railway Noise, Acoustics |
|
Kyohei NAGAO Assistant Senior Researcher, Current Collection Laboratory, Railway Dynamics Division Research Areas: Catenary-Pantograph Interaction, Dynamics of Machinery, Mechanical Vibration |
|
Isamu MAKARA Assistant Senior Researcher, Current Collection Laboratory, Railway Dynamics Division (Former) Research Areas: Aerodynamics of Pantograph, Catenary-Pantograph Interaction |
|
Yusuke WAKABAYASHI, Ph.D. Chief Researcher, East Japan Railway Company, Next-generation Rolling Stock System Unit, Research and Development Center of JR East Group Research Areas: Aerodynamic Noise Reduction |
