Research Article  Open Access
Shaking Table Array Tests of an UltraHighVoltage CupType Transmission TowerLine System
Abstract
Ultrahighvoltage (UHV) cuptype transmission towers supported with longspan transmission lines are unavoidably subjected to the coupling action between the towers and the transmission lines. Therefore, investigating how towerline coupling affects UHV cuptype transmission towers is important. In this study, three shaking table array tests of an UHV cuptype transmission towerline system were carried out to investigate the dynamic characteristics of the coupling action between the towers and transmission lines based on the following four comparative models: a singletower model, a singletower model with suspended lumped masses, a threetowertwoline model, and a fivetowerfourline model. The test results demonstrated that the towerline coupling interaction had a significant effect on the dynamic characteristics and seismic responses, as the suspended conductor line and the suspended lumped mass decreased the frequency of the transmission tower. Under longitudinal ground motion, the model with the suspended lumped mass had the lowest peak acceleration response and the largest peak displacement response. Under the same ground motion, the four models had similar peak strains in the longitudinal direction. Under transversetheline ground motion, the model with the suspended lumped mass had the lowest peak acceleration response and the smallest peak responses for displacement and strain in the transverse direction; therefore, this model is inappropriate for the simulation and seismic evaluation of transmission towerline systems.
1. Introduction
Electricity transmission systems consisting of transmission towers and transmission lines have been widely recognized as lifeline systems; in the system, the transmission tower is a type of lattice tower that supports the overhead transmission lines. Compared with other power equipment, the ultrahigh voltage (UHV) transmission tower structures are characterized by great heights, long spans, and complex towerline coupling effects. However, this type of structure is generally considered to be less impacted by seismic action, and relevant studies are rare [1–4]. The mechanical performance of this type of structure under disastrous dynamic wind and ice loads is the main focus of the research performed worldwide [5–9].
However, earthquakes have frequently occurred in recent years worldwide, and the power transmission systems have been severely damaged in earthquakeridden regions [10]. The earthquake damage investigations indicate that transmission line damage is severe [11]. As revealed by the investigations of the damage to transmission lines caused by several typical major earthquakes, seismic damage mainly manifests as tower tilting, insulator breakage, and foundation settling and dislocation. For example, in the 1994 Northridge earthquake in the United States, a few highvoltage transmission towers were damaged; the level of damage to the transmission towers was related to the intensity of the ground motion and the seismic design. In the 1995 Hanshin earthquake in Japan, 23 overhead transmission lines were damaged and 11 steel transmission towers collapsed. During the 1999 ChiChi earthquake, transmission lines were severely damaged, including 28 345kilovolt UHV transmission lines, amounting to a damage rate close to 30%. In the 2008 Wenchuan earthquake, more than 20 over10kilovolt transmission towers collapsed, and approximately 16 towers were locally damaged, resulting in severe damage[12]. During the 2013 Lushan earthquake in Sichuan, 265 over10kilovolt transmission towers with distribution lines outofservice and 224 over35kilovolt lines were severely damaged, reaching a damage rate of 50% and resulting in huge losses [13]. Considering the harsh and complex geological conditions of the areas along UHV transmission lines, if one tower is damaged or collapses under a strong earthquake, the operation of the whole transmission line will be affected. Therefore, the mechanical mechanism and failure mode of transmission tower structures under strong earthquakes and the corresponding seismic theory and methods have garnered increasing amounts of attention [14, 15].
Shaking table tests can reproduce the actual stress state of a structure under an earthquake well and can be used to study the structural dynamic performance, structural response, and failure mode under seismic action. However, the UHV transmission towerline system is characterized by tall towers (the UHV transmission towers built in China generally exceed 100 m in height) and extremely large spans (e.g., longer than 1000 m), so it is difficult to meet the experimental requirements with single shaking table test equipment. Therefore, shaking table tests of UHV transmission towerline systems have not often been carried out as conventional civil structures. The earliest known shaking table tests of highvoltage transmission towers were conducted in the 1980s. Kotsubo et al. [16] conducted a shaking table test of three transmission towers in south eastern Kyushu, Japan. The researchers considered the influence of the conductor lines and the ground lines on the adjacent towers, obtained the dynamic characteristics of the transmission towers under earthquakes, and proposed a corresponding seismic calculation method for the transmission towers. Filiatrault [17] conducted a shaking table test of five different electrical substation equipment components, considered the interactions between the conductor lines and insulators with different degrees of relaxation under earthquakes, and concluded that the three flexible conductor lines with different degrees of relaxation had little impact on the insulators under earthquakes. Kempner et al. [18] conducted dynamic tests and theoretical analysis of the modeling of guyed towers. Tian et al. [19] considered the length limitation of the shaking table for the extremely large span of a highvoltage transmission towerline system and noted that the towertoline scale factor for a highvoltage transmission towerline system is difficult to maintain. Based on numerical analysis, those of the conductor line and ground line were corrected with the similitude ratio of the transmission tower remaining constant. Shi et al. [20] conducted a shaking table test of a simplified longspan transmission towerline coupling system using steel rods with a diameter of 10 mm to simulate the transmission tower and a steel chain to simulate the conductor line. Three steel rods were connected by two layers of steel chains to form a transmission towerline coupling system model. This test verified the accuracy of the proposed theoretical calculation method for the transmission lineline coupling system and reflected the dynamic characteristics of a highvoltage transmission towerline system. However, the experimental model was overly simplified, and the chains that were used to simulate the conductor lines were relaxed and failed to truly reflect the mechanical performance of the tower structure and conductor line. Considering the deficiencies of the shaking table tests by Bai et al., [21, 22] proposed a new model design method for shaking table tests for a UHV transmission towerline coupling system with double circuits on the same tower and geometric nonlinearity and conducted a shaking table test of a UHV grid transmission towerline system. Aiming at a coupling system consisting of two different structural forms, i.e., a highly geometrically nonlinear conductor line and ground line and an elastic transmission tower structure, a design method for a separate model with dynamic parameters was proposed in the experimental scheme to address the problem that a towerline system with a uniform scale cannot meet the length limitation of a shaking table. Xie et al. [23–25] conducted a number of shaking table tests of a highvoltage transmission anglesteel towerline system and obtained the dynamic response characteristics of UHV angle steel transmission towers, the effects of the mass of the conductor line and ground line on the dynamic response of the tower structure, and the nonlinear vibration responses of the conductor line and ground line of a UHV transmission towerline coupling system. An improved test method was proposed, and many important conclusions were drawn, which provided a reference for future shaking table tests of towerline system models. However, most of the research results were obtained based on single shaking tables.
The above studies show that the main problem in the shaking table tests of UHV transmission towerline systems are the length limitations of single shaking tables; the use of an array of shaking tables to capture the interactions in transmission towerline systems is scarce, making it difficult to consider the effect of towerline coupling on the transmission towers’ responses. In this paper, an array of three shaking tables are used to test UHV cuptype transmission towersline systems to investigate the interactions in transmission towerline systems.
2. Model Design of the UHV CupType Transmission TowerLine System
A typical transmission tower prototype, the ZBC30102Atype selfsupporting suspension tower in the XimengShandong UHV project, is chosen for the shaking table tests. The tower has a height of 73.8 m and a line span of 500 m; the prototype has three conductor lines, which are suspended in the middle and on both sides of the tower head through insulators. The conductor line and the ground line types are JL/G1A630/45 and JLB20A170, respectively. The transmission tower prototype is shown in Figure 1.
2.1. Design and Manufacture of the Tower
2.1.1. Geometric Similitude Ratio and Material Selection
Considering the size limitations of the shaking tables, which have side lengths of 2.5 m, the geometric similitude ratio of the transmission tower is assumed to be 1 : 15.
If all the angle steel members are scaled based on the geometric scale ratio, for the cross section, it is difficult to make the members because their thickness would only be 0.02 mm. Therefore, the transmission tower model uses four specifications of the angle steel (L30 × 3, L25 × 3, L20 × 3, and L10 × 1). The L30 angle steel is used for the main members and diagonal members of the tower leg and tower body, the L25 angle steel is used for the main members and diagonal members of the upper and lower cranks, and the L20 angle steel is used for main members and diagonal members of the ground line supports and crossarms. The auxiliary members of the model tower are made of L10 angle steel, and welded connections are used.
2.1.2. Other Similitude Coefficients
The same material is chosen to use in the towers of the experimental model as is used in the prototype; therefore, the similitude ratio for the elastic modulus E is 1 : 1, and considering the noise and bearing capacity of the shaking tables, the similitude coefficient of acceleration is assumed to be 3 : 1. Then, other similitude coefficients are derived as the preliminary similarity relationships by using dimensional analysis and based on the Buckingham π theorem for the transmission tower model design (Table 1).

The total mass of the prototype transmission tower is 83544.6 kg, and the total mass of the scaled transmission tower model is 642.65 kg.
2.1.3. The Counterweight of the Model
Because the density similitude ratio is 5 : 1, the total mass of the scaled transmission tower model is 642.65 kg, the weight of the transmission tower without a counterweight is 128.53 kg, and the mass of the actual scaled transmission tower without a counterweight is 129 kg; to evenly apply the counterweight, the actual balancing weight is 555 kg. The actual total mass of the manufactured transmission tower model is 684 kg, and the actual density similitude ratio is 5.3 : 1. In the tests, two counterweights are processed, with masses of 2 kg and 0.25 kg. The counterweights are shown in Figure 2.
(a)
(b)
2.2. Design and Manufacture of the Conductor Line and Ground Line
2.2.1. Similitude Coefficients of the Conductor Line and Ground Line
The tests were carried out using the threeshakingtable array system at Fuzhou University. A 4.1 m × 4.1 m horizontal bidirectional shaking table is in the middle, and two 2.5 m × 2.5 m horizontal bidirectional shaking tables are on each side. The maximum distance between the middle table and the side tables is 10 m, and the horizontal span of the prototype model is 500 m. If the tests are based on a geometric scale ratio of 1 : 15, the dimensions of the shaking tables are unable to meet the requirements. Therefore, in this study, the similitude ratio of the dynamic characteristics and the similitude ratio of the inertial mass are assumed separately to design the similitude coefficients of the conductor lines and ground lines.
(1) Design of the Frequency Similitude Ratio of the Conductor Line and Ground Line. When designing the model of the transmission towerline coupling system, the similitude ratio of the transmission tower should be close to that of the conductor line and ground line.
The fundamental natural frequency of the conductor line and ground line is as follows:where is the fundamental natural frequency of the conductor line and ground line, is the gravity acceleration constant, and is the sag in the conductor line and ground line.
The relationship between the frequency similitude ratio and sag similitude ratio of the conductor line and ground line can be derived as follows:
The transmission tower has a frequency similitude ratio of 6.7 : 1, and the conductor line and ground line have a frequency similitude ratio of 6.7 : 1 and a sag similitude ratio of 44.9 : 1.
(2) Design of Mass Similitude Ratio of the Conductor Line and Ground Line. If the geometric similitude ratio of the conductor line and ground line is assumed to be the same as that of the transmission tower (1 : 15), since the span of the prototype tower is 500 m, the distance between the model towers would be 33.33 m, exceeding the maximum distance (10 m) between the middle table and the side table and, hence, not meeting the requirements. Therefore, a span similitude ratio of 1 : 55 is selected for the actual tests, and accordingly, the actual distance between the middle tower and the side tower is 9.09 m.
The mass similitude ratio of the conductor line and ground line is the same as that of the transmission tower (1 : 130). The total mass of the prototype conductor line is 2.06 × 500 = 1030 kg, so the mass of the model conductor line is 7.92 kg. The total mass of the prototype ground line is 0.49 × 500 = 245 kg. The actual linear densities of the model conductor line and ground line are calculated to be 0.88 kg/m and 0.49 kg/m, respectively (Table 2).

2.2.2. Manufacture of Conductor Line and Ground Line
Steel wires with the diameters of 4 mm and 2 mm are used for the conductor line and ground line, respectively, in the experimental model. The physical components used for the conductor line and ground line are shown in Figure 3.
(a)
(b)
According to the mass similitude ratio of the conductor line and ground line models, since the density of the steel wire is lower than that of the design model, counterweights are required. The counterweights of the conductor line and ground line are 8 kg and 4 kg, respectively. The counterweight blocks of the corresponding masses are placed uniformly along the conductor line and ground line.
2.3. Design and Manufacture of the Insulator
The Vshaped insulator is simulated by a galvanized plain round steel bar with a diameter of 8 mm and a length of 750 mm, and the Ishaped insulator is simulated by a rigid steel plate with a length of 600 mm and a width of 50 mm. The physical components used for the insulator models are shown in Figure 4. To ensure that the insulator is hinged at two ends with the conductor line, a Ushaped steel wire lock is selected to simulate the suspension clamp in the prototype tower, as shown in Figure 5.
(a)
(b)
2.4. Equivalent Tower
The equivalent tower is the side tower in the threetowertwoline model and the fivetowerfourline model. The equivalent tower is used to simulate the boundary conditions without considering the similarity relationship with the tower model. In the tests, two equivalent towers are manufactured using welded L50 × 4, as shown in Figure 6. The function of the equivalent tower is similar to that of a corner tower in practical engineering projects.
2.5. Test Models
2.5.1. SingleTower Model
The test tower is anchored to the rigid base of the shaking table.
2.5.2. ThreeTowerTwoLine Model
The threetowertwoline model is a singletowertwoline model that has a cuptype tower as the middle tower and equivalent towers as the side towers. All three towers are anchored to the rigid base of the shaking table. The conductor line and ground line are fixed on the equivalent towers. The boundary conditions are shown in Figure 7, and the physical components used for the threetowertwoline model are shown in Figure 8.
2.5.3. FiveTowerFourLine Model
The fivetowerfourline model is an improved version of the threetower and twoline model. That is, two new tower models, which are the same as the middle tower, are placed at the locations of the original equivalent towers, which are relocated to the outer side of the new side main towers, along the line direction, and the boundary conditions are simulated through the connection between the equivalent towers and conductor lines, as shown in Figure 9. The physical components used for the fivetowerfourline model are shown in Figure 10.
3. Shaking Table Test Setup
3.1. Selection of Ground Motions
Considering the long span of transmission lines and the different site conditions in the regions that the lines traverse, the El Centro record, Taft record, and artificial waves are selected for site classes II, III, and IV, respectively. The duration of the ground motion is compressed to 1/6.7 of the original, based on the similarity relationship. The original wave is compressed according to the time or frequency similitude ratio and is then applied to the seismic excitation tests of the scale models. The 0.07 g seismic records are shown in Figure 11.
(a)
(b)
(c)
3.2. Sensor Layout
Two types of acceleration sensors, namely, a piezoresistive accelerometer and piezoelectric accelerometer, are used in the tests. The piezoresistive accelerometer has a large measuring range and is mainly used to measure the seismic responses of the top of the tower and the conductor and ground lines. The piezoelectric accelerometer has a slightly smaller measuring range and is mainly used to measure the seismic responses of the tower leg and body. Two types of displacement sensors, namely, a laser displacement sensor and guyed displacement sensor, are used in the tests. Resistive strain gauges are used in the tests to test the strain of the tower model members.
3.2.1. Sensor Layout for the SingleTower Model
In the singletower model tests, a total of 22 acceleration sensors, 25 strain gauges, and 14 displacement sensors are deployed. The detailed placements of the sensors are shown in Figures 12–14, in which “X” indicates the longitudinal direction, “Y” indicates the transverse direction, as shown by the coordinates in the lower left corner, “A” denotes an acceleration sensor, “S” denotes a strain gauge, and “D” denotes a displacement sensor.
3.2.2. Sensor Layout for the ThreeTowerTwoLine Model
(1) Acceleration Sensors. The layout of the acceleration sensors on the threetowertwoline model is exactly the same as that on the body of the singletower model with counterweights. A total of 16 acceleration sensors are deployed on the conductor lines. A total of six acceleration sensors are placed on the ground line. The acceleration sensors are evenly distributed along the conductor and ground lines. The layout of the acceleration sensors on the conductor and ground lines is shown in Figure 15, where C represents the acceleration sensor and the numeral represents the numbering of the sensors.
In the “threetowertwoline” model, four acceleration sensors are placed on each of the two equivalent towers. Specifically, for each equivalent tower, one acceleration sensor is placed in the X and Y directions on top of the tower, and one acceleration sensor is placed in the X and Y directions on the base of table. Thus, a total of eight acceleration sensors are placed. The detailed layout is shown in Figure 16, where “L” is the left equivalent tower, “R” is the right equivalent tower, “A” is the acceleration, “S” is the strain, “D” is the tower top, “T” is the tower leg, “M” is the table top, and “X” and “Y” are the two vibration directions. The acceleration sensors are placed at the middle point on the top of the equivalent tower, and the strain gages are placed on the tower legs along the line direction.
(2) Displacement Sensors. The layout of the displacement sensors on the “threetowertwoline” model is the same as that on the body of the singletower model with counterweights. Thus, the detailed layout is the layout of the displacement sensors on the singletower model (Figure 13).
(3) Strain Gauge Layout. The layout of the strain gauges on the “threetowertwoline” model is the same as that on the body of the singletower model. The detailed layout is shown in Figure 14, where “S” represents a strain gauge and the numeral represents the numbering of the gauges.
3.2.3. Sensor Layout for the FiveTowerFourLine Model
(1) Acceleration Sensor. There are three model towers in the “fivetowerfourline” model. Due to the limitations of the sensors and acquisition channels, in particular, 18 acceleration sensors are deployed in the middle tower. The specific layout is illustrated in Figure 17(a), where A represents an acceleration sensor and the numeral represents the numbering of the sensors.
In the fivetowerfourline model, the layout of the sensors on the two model towers on the left and right small tables reference that on the middle tower. A total of 14 acceleration sensors are placed on the right main tower, as shown in Figure 17(b), where R represents the right side, A represents an acceleration sensor, and the numeral represents the numbering of the sensors consistent with that of the middle main tower, indicating that the position of a sensor on the side main tower corresponds with that on the middle tower. A total of 11 acceleration sensors are placed on the left main tower, as shown in Figure 17(c), where L represents the left side, A represents an acceleration sensor, and the numeral represents the numbering of the sensors consistent with that of the middle main tower, indicating that the position of a sensor on the side main tower corresponds to that on the middle tower.
The layout of the sensors on the equivalent tower in the fivetowerfourline model is identical to that on the equivalent tower of the threetowertwoline model. The specific layout is illustrated in Figure 16.
(a)
(b)
(c)
A total of three acceleration sensors are deployed on the conductor line, and one acceleration sensor is deployed on the ground line. The specific sensor layout is shown in Figure 18, where C represents an acceleration sensor, the numeral represents the numbering of the sensors, and the arrow represents the direction of the vibration.
(2) Displacement Sensors. The layout of the displacement sensors on the towers of the fivetowerfourline model is identical to that on the tower of the singletower model with counterweights. The specific layout references the layout of the displacement sensors on the singletower model (Figure 13), where D represents the displacement and the numeral represents the numbering of the sensors.
In the fivetowerfourline model, the layouts of the sensors on the two model towers on the left and right small tables reference that on the middle tower, but the sensors are all deployed above the tower structure, where the layout of the sensors above the structure of the right main tower is consistent with that of the middle tower, with 8 displacement sensors placed (Figure 19(a)), while the sensors are symmetrically deployed only on the upper structure of the left main tower, with 4 displacement sensors placed (Figure 19(b)). In the figure, L stands for the left side, R is the right side, D is the displacement sensor, and the numeral is the numbering of the sensors. When the numeral of the main side tower is consistent with that of the middle tower, the location of the sensor on the main side tower corresponds with that on the middle tower. In addition, the direction of the sensor is represented by X or Y. The specific layout is illustrated in Figure 19.
(a)
(b)
The fivetowerfourline model has a total of 73 strain gauges, including 28 on the middle tower, 23 on the right main tower, 22 on the left main tower, and 2 on the equivalent tower. Figure 20 illustrates the specific layout of the strain gages.
(a)
(b)
(c)
3.3. Test Program
The basic dynamic characteristics tests and seismic timehistory response tests are conducted on the singletower model, the singletower model with suspended lumped masses, the threetowertwoline model, and the fivetowerfourline model. A total of 21 runs were performed, as listed in Table 3.

4. Test Results
To study the influence of the coupling effect of the transmission towerline system on the transmission tower, dynamic characteristics tests and the shaking table tests are performed on the singletower model, the singletower model with suspended lumped masses, the threetowertwoline model, and the fivetowerfourline model. The results of the frequency, damping, acceleration, displacement, and strain of the middle tower are obtained from the tests.
4.1. Dynamic Characteristics Test Results
Figure 21 shows the spectral analysis results of the acceleration timehistory curve at the midpoint of the middle tower model head in the transverse direction (firstorder) and the longitudinal direction (secondorder) of the four models. Table 4 compares the basic dynamic characteristics of the middle towers of the four models. The results show that the suspension of the conductor lines and lumped masses decreases the frequency of the middle tower. Compared with the singletower model, the firstorder frequency of the single tower with suspended lumped masses model decreases by 5.08%, and the secondorder frequency decreases by 4.03%. The frequencies of the two models with suspended conductor lines are very close. Compared with those of the singletower model, the firstorder frequency of the model with suspended conductor lines decreases by 0.82%, and the secondorder frequency decreases by 4.0%. In terms of structural damping, the suspension of clumped masses and conductor lines increase the damping ratio of the structure, with the damping ratio of the threetowertwoline model reaching a maximum of 2.99%. These results show that there is no significant difference regarding the effect of towerline coupling for the dynamic characteristics between the threetowertwoline model and the fivetowerfourline model. The effect of towerline coupling decreases the frequency of the tower and increases the damping ratio. The singletower model with suspended lumped masses can simulate the effect of towerline coupling on the longitudinal dynamic characteristics of the transmission tower well, but this model faces difficulty in satisfactorily simulating the effect of towerline coupling on the transverse characteristics of the transmission tower.
(a)
(b)

4.2. Comparative Analysis of the Dynamic Results of the Middle Towers from the Shaking Table Tests
(1) Comparative Analysis of the Acceleration Response. Figure 22 shows the longitudinal peak acceleration responses under the unidirectional input of a 0.21 g artificial wave. The results show that the peak acceleration response curves of the tower accelerations of the different models basically coincide, and the trends are similar. The peak acceleration response gradually increases from the base of the tower to the top, reaching the maximum at the top of the tower.
Table 5 shows the longitudinal peak acceleration responses of the middle towers of the four models under each wave. The results show that the peak acceleration response is the largest under the artificial wave and the smallest under the El Centro wave. The model with the suspended lumped mass has the smallest acceleration response. Therefore, the use of a singletower model with suspended lumped masses to simulate a threetowertwoline model or a fivetowerfourline model is inaccurate. Except for the 0.63 g test condition, the effect of the towerline coupling increases the longitudinal peak acceleration response of the tower.

Figure 23 shows the transverse peak acceleration response under the unidirectional input of a 0.21 g artificial wave. The results show that except for the single tower with suspended lumped masses model, the peak acceleration response curves of the towers of the other models basically coincide and have similar trends. The comparison shows that the peak accelerations are relatively close within the lower portion of the tower structure, while the peak acceleration responses at the crossarm vary noticeably. The peak acceleration of the singletower model with suspended lumped masses is the smallest at the crossarm, and the peak acceleration of the singletower model is the largest at the crossarm.
Table 6 shows the transverse peak accelerations of the middle towers of the four models under each wave. The results show that the peak acceleration response is the largest under the artificial wave and is the smallest under the El Centro wave. The model with suspended lumped masses has the lowest peak acceleration response, and the singletower model with counterweights has the highest peak acceleration response. Therefore, the suspended masses and conductor lines are decreasing the transverse seismic response of the model tower. In comparison, the single tower with suspended lumped masses tested using a single shaking table has the lowest seismic response, so the use of this model to simulate the towerline system is inaccurate. Under the three seismic conditions of the same seismic intensity level, the average transverse peak acceleration response of the threetowertwoline model is greater than that of the fivetowerfourline model; the effect of towerline coupling decreases the transverse peak acceleration response of the tower.

(2) Comparative Analysis of the Displacement Responses. Figure 24 shows the longitudinal peak displacement responses under the unidirectional input of a 0.21 g artificial wave. The results show that except for that of the singletower model with suspended lumped masses, the peak displacement response gradually increases from the base of the tower to the top, reaching the maximum peak displacement at the top of the tower.
Table 7 shows the longitudinal peak displacements of the main towers of the four models under each wave. The results show that the peak displacement response is the largest under the artificial wave and is the smallest under the El Centro wave. The singletower model with suspended lumped masses has the maximum displacement response. Under the three test conditions at the same seismic level, the average of the longitudinal peak displacement responses of the threetowerconsistency model is larger than that of the fivetowerconsistency model. Except for 0.63 g test condition, the effect of towerline coupling increases the longitudinal peak displacement response of the tower.

Figure 25 shows the transverse peak displacement responses under the unidirectional input of a 0.21 g artificial wave. The results show that the trends of the peak displacement response curves of the towers are similar. The peak displacement response gradually increases from the base of the tower to the top, reaching the largest peak displacement at the top of the tower.
Table 8 shows the transverse peak displacements of the main towers of the four models under each wave. The results show that the peak displacement response is the largest under the artificial wave and is the smallest under the El Centro wave. The singletower model with suspended lumped masses has the smallest peak displacement response, and the threetowertwoline model has the largest peak displacement response. Therefore, the use of the singletower model with suspended lumped masses to simulate the towerline system is inaccurate. Under the three test conditions at the same seismic level, the average of the transverse peak displacement response of the threetowertwoline model is larger than that of the fivetowerfourline model; the effect of the threetowertwoline model’s towerline coupling increases the transverse peak displacement response of the tower, while the effect of the fivetowerfourline model’s towerline coupling decreases the transverse peak displacement response of the tower.

(3) Comparative Analysis of the Strain Responses. Figure 26 shows that for the four models under a 0.21 g unidirectional seismic input, the longitudinal peak strain variation curves of the towers under the crank arms are consistent in both shape and magnitude, and the strain is relatively large at the locations of the tower head, crank arm, and tower legs.
Table 9 shows the longitudinal peak strain values of the main tower strains of the four models under each wave. The results show that the peak strain response is the largest under the artificial wave and is the smallest under the El Centro wave. The longitudinal peak strains of the main towers of the four models are relatively close under the same wave. Except for the 0.63 g test condition, the effect of towerline coupling increases the longitudinal peak strain response of the tower.

Figure 27 shows the curves of the transverse peak strain values of the tower body structure for the four models under a 0.21 g unidirectional seismic input. The curves are consistent in shape, with parallel decreasing trends. The largest peak strain occurs in the singletower model, reaching approximately 200 με, and the smallest peak strain occurs in the singletower model with suspended lumped masses, which is approximately 1/2 of that of singletower model without suspended lumped masses.
Table 10 shows the transverse peak strain values of the main tower of the four models under each wave. The results show that the peak strain response is the largest under the artificial wave and the smallest under the El Centro wave. Except for the 0.21g test condition, the model with suspended lumped masses has the smallest peak strain response, and the singletower model has the largest peak strain response. Therefore, the suspended masses and conductor lines reduce the transverse seismic response of the model tower. The effect of towerline coupling decreases the transverse peak strain response of the tower.

5. Conclusions
In this study, threeshakingtable array system tests of UHV cuptype transmission towerline systems were carried out, and the dynamic characteristics and seismic responses of a singletower model, a singletower model with suspended lumped masses, a threetowertwoline model, and a fivetowerfourline model were studied. The test results demonstrated the following:(1)The suspended conductor lines and suspended lumped masses reduce the frequency of the middle tower. The singletower model with suspended lumped masses can simulate the effect of towerline coupling on the longitudinal dynamic characteristics of the transmission tower well but cannot simulate the transverse dynamic characteristics of the transmission tower. The dynamic characteristics of the threetowertwoline model are very close to those of the fivetowerfourline model. The effect of towerline coupling decreased the frequency of the tower but increased the damping ratio.(2)Under longitudinal ground motion, except for the singletower model with suspended lumped masses, the peak accelerations and peak displacements of the other models gradually increase from the base of the tower to the top, reaching the highest peak acceleration and peak displacement at the top of the tower. The peak strain curves of the four models below the crank arm are consistent in shape and magnitude, and the strains are relatively large at the tower head, crank arms, and tower legs. The model with suspended lumped masses has the lowest peak acceleration response and the largest peak displacement response. The longitudinal peak strains of the main towers of the four models under waves are relatively close. Except for 0.63 g seismic condition, the effect of towerline coupling increased the longitudinal peak acceleration, displacement, and strain response of the tower.(3)Under transverse ground motion, except for the singletower model with suspended lumped masses, the peak acceleration responses and peak displacements of the other models gradually increase from the base of the tower to the top, reaching the highest peak acceleration and peak displacement values at the top of the tower. The transverse peak strain curves of the tower structure are consistent in shape, with a parallel decreasing trend, and the model with counterweights has the largest peak strain. The model with suspended lumped masses has the smallest transverse peak acceleration responses for acceleration, displacement, and strain, and use of this model to simulate the towerline system is inaccurate. Under the three seismic conditions of the same seismic intensity level, the effect of towerline coupling decreases the transverse peak acceleration response of the tower, the effect of the threetowertwoline model’s towerline coupling increases the transverse peak displacement response of the tower, while the effect of the fivetowerfourline model’s towerline coupling decreases the transverse peak displacement response of the tower. Except for 0.21 g seismic condition, the effect of towerline coupling decreased the transverse peak strain response of the tower.
Data Availability
The data used to support the findings of this study are included within the article.
Conflicts of Interest
The authors declare that there are no conflicts of interest regarding the publication of this paper.
Acknowledgments
This study was supported by a fund from the Scientific Research Fund of the Institute of Engineering Mechanics, CEA (Nos. 2019B05 and 2019C10), the National Natural Science Foundation of China (Nos. 51878631 and 51608287), and the National Key Research and Development Program of China (Nos. 2017YFC1500605 and 2018YFC150460201).
References
 L. Tian, R. S. Ma, F. Zhu et al., “Seismic research progress of high voltage transmission towerline system,” World Earthquake Engineering, vol. 30, no. 1, pp. 87–95, 2014. View at: Google Scholar
 H. N. Li and H. F. Bai, “Stateoftheart review on studies of disaster resistance of highvoltage transmission towerline systems,” China Civil Engineering Journal, vol. 40, no. 2, pp. 39–46, 2007, in Chinese. View at: Google Scholar
 L. Tian, W. F. Li, and Z. L. Wang, “Study on suspended mass pendulums for vibration control of transmission tower under seismic excitation,” Advanced Materials Research, vol. 446–449, pp. 3323–3327, 2012. View at: Publisher Site  Google Scholar
 H.N. Li, W.L. Shi, G.X. Wang, and L.G. Jia, “Simplified models and experimental verification for coupled transmission towerline system to seismic excitations,” Journal of Sound and Vibration, vol. 286, no. 3, pp. 569–585, 2005. View at: Publisher Site  Google Scholar
 S. C. Yang and H. P. Hong, “Nonlinear inelastic responses of transmission towerline system under downburst wind,” Engineering Structures, vol. 123, no. 15, pp. 490–500, 2016. View at: Publisher Site  Google Scholar
 F. Yang, H. Dang, H. Niu, H. Zhang, and B. Zhu, “Wind tunnel tests on wind loads acting on an angled steel triangular transmission tower,” Journal of Wind Engineering and Industrial Aerodynamics, vol. 156, pp. 93–103, 2016. View at: Publisher Site  Google Scholar
 F. L. Yang, H. J. Zhang, J. B. Yang, H. X. Dang, and J. J. Liu, “Bearing capacity analysis and load values of transmission towers under thunderstorm downburst,” Proceedings of the CSEE, vol. 34, no. 24, pp. 4179–4186, 2014, in Chinese. View at: Google Scholar
 W. J. Lou, X. Liang, Y. Jiang, X. H. Jin, and Z. H. Wang, “Windinduced response and wind load factor of transmission tower under terrain B wind field and typhoon wind field,” Journal of Vibration and Shock, vol. 32, no. 6, pp. 13–17, 2013. View at: Google Scholar
 J. Li, Q. Yan, Q. Xie, and J. Chen, “Wind field measurements and windinduced vibration responses of transmission tower during typhoon wipha,” Journal of Architecture and Civil Engineering, vol. 26, no. 2, pp. 1–8, 2009. View at: Google Scholar
 Z. Yan, B. Z. Zhu, X. F. Zi, and Y. X. Xi, “Seismic response analysis of towerfoundationsoil dynamic interaction in transmission line engineering,” Advanced Materials Research, vol. 1065–1069, pp. 1112–1116, 2014. View at: Publisher Site  Google Scholar
 Q. Xie, “Stateoftheart of seismic disaster research and emergency response of electric power system,” Electric Power Construction, vol. 29, no. 8, pp. 2–6, 2008. View at: Google Scholar
 Z. Y. Zhang, B. Zhao, W. W. Cao, Y. Li, and Z. X. Ge, “Investigation and preliminary analysis of damages on the power grid in the wenchuan earthquake of M8.0,” Electric Power Technologic Economics, vol. 20, no. 4, pp. 1–4, 2008. View at: Google Scholar
 H. B. You and F. X. Zhao, “M7.0 earthquake in lushan and damage cause analysis of power facilities,” Electric Power Construction, vol. 34, no. 8, pp. 100–104, 2013, in Chinese. View at: Google Scholar
 Z. Xu, W. M. Wang, Q. H. Lin et al., “Progressive collapse analysis of a simplified model for transmission towerline coupled system under seismic action,” China Earthquake Engineering Journal, vol. 7, no. 2, pp. 304–309, 2015, in Chinese. View at: Google Scholar
 Q. H. Lin, W. M. Wang, D. J. Cao, S. Jin, L. Tian, and L. Wang, “Progressive collapse analysis of highvoltage transmission tower under earthquake action,” Sichuan Building Science, vol. 40, no. 3, pp. 157–161, 2014. View at: Google Scholar
 S. Kotsubo, T. Takanishi, and T. Sonoda, “Dynamic tests and seismic analysis of high towers of electrical transmission line,” Transactions of the Japan Society of Civil Engineers, vol. 15, pp. 72–75, 1985. View at: Google Scholar
 A. Filiatrault and C. Stearns, “Seismic response of electrical substation equipment interconnected by flexible conductors,” Journal of Structural Engineering, vol. 130, no. 5, pp. 769–778, 2004. View at: Publisher Site  Google Scholar
 J. L. Kempner and R. C. Stroud, “Transmission line dynamic/static structural testing,” Journal of the Structural Division, vol. 107, no. 10, pp. 1895–1906, 1981. View at: Google Scholar
 L. Tian, H. N. Li, and L. Z. Huang, “Nonproportional problem in shaking table model of a power transmission towerline system,” Journal of Vibration and Shock, vol. 31, no. 6, pp. 114–118, 2012, in Chinese. View at: Google Scholar
 W. L. Shi, L. H. Nan, and J. L. Guang, “Shaking table test of coupled system of transmission lines and towers,” Engineering Mechanics, vol. 23, no. 5, pp. 89–93, 2006, in Chinese. View at: Google Scholar
 J. Bai, Q. Xie, and S. T. Xue, “Design theory and method of shaking table test model for a UHV double circuit transmission towerline coupled system,” Journal of Vibration and Shock, vol. 32, no. 22, pp. 47–52, 2013. View at: Google Scholar
 J. Bai, Q. Xie, S. T. Xue, and D. M. Shen, “Shaking table test model design for UHV latticed transmission towerline coupling system,” Journal of Hunan University (Natural Science Edition), vol. 40, no. 4, pp. 20–25, 2013. View at: Google Scholar
 Q. Xie, J. Bai, S. T. Xue, and Q. Q. Zhi, “Shaking table test on nonlinear vibration responses of UHV transmission towerline coupling system,” Earthquake Engineering and Engineering Vibration, vol. 32, no. 5, pp. 103–110, 2012. View at: Google Scholar
 Q. Xie, J. Bai, S. T. Xue, and Q. Q. Zhi, “Shaking table test on 1000 kV angle steel transmission tower,” Power System Technology, vol. 37, no. 1, pp. 58–64, 2013, in Chinese. View at: Google Scholar
 J. Bai, Q. Xie, and S. T. Xue, “Shaking table tests on 1000 kV UHV AC double circuit transmission towerconductor coupling systems,” Proceedings of the CSEE, vol. 33, no. 7, pp. 116–123, 2013. View at: Google Scholar
Copyright
Copyright © 2019 Fei Wang et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.