This study aimed to analyze dynamic performance of the low- and medium-speed maglev train on the turnout, by conducting a series of field tests on the Fenghuang maglev sightseeing express line. This study systematically analyzed and evaluated dynamic responses of vehicle body, suspension bogie, short rail joint, motor-driven long-span girder (LSG) and continuous beam of the turnout at different speeds and loads in time and frequency domains. Test results obtained in this study indicated that: 1) speed and load of the maglev train affected acceleration amplitude of the train and turnout; 2) vertical sperling indexes of the test train were less than 2.5, indicating that the train had good ride quality; and 3) resonance phenomenon of track occurred at 40 km/h. The findings of this study may serve as references for model validation and optimized design of low- and medium-speed maglev train.
Guided transport has been developed fast all over the world in the past decades due to the higher transport capacity and energy efficiency in operation compared with road traffic and air traffic. Nevertheless, conventional rail transport systems are limited by a series of technical issues related to adhesion, rail-wheel wear, noise and vibration . Maglev systems come as new types of guided transport systems which have the advantages of non-contact operation, low vibration, low noise, no risk of derailment, flexible route selection and low comprehensive construction costs and thus has received widespread attention for its development and application in Japan, Korea and China . Especially, low- and medium-speed maglev lines have been commercially operated, such as China’s Changsha Airport Line (18.55 km), Japan’s Linimo (8.9 km) and South Korea’s Incheon Airport Maglev Line (6.1 km) .
With the operation of the maglev line, a strong coupling interaction has been shown to occur between the maglev train, suspension control and guideway systems, thus comprising a time-varying system problem. Related research has been conducted to analyze the dynamic performance of low- and medium-speed maglev system , , , , . The interaction effect contributes an additional damping or mass effect to the vibration of the guideway, which causes temporal variations in the natural frequency and modal damping ratio of the guideway. The flexibility of the guideway, which acts as an external force of excitation on the train, may also generate dynamic instability and unacceptable vibration, such as self-excited and resonant vibrations , . As the frequency of the external excitation acting on the train approaches the natural frequency of the train vehicles, the dynamic response of the train tends to increase significantly, resulting in resonance , . Resonance-induced instability of the levitation gap, which can lead to failure of suspension control, has been observed in maglev trains on both test and commercial lines. Hence, to ensure the operational stability and safety of maglev trains, it is essential to study the complicated dynamic interaction mechanism of the maglev train-suspension-guideway system.
The dynamic interaction of the maglev train-suspension-guideway system is relatively difficult to evaluate because it involves mechanics, electronics and control theory . Early studies used simplified analytical models to examine the dynamic interaction of the maglev system. The guideway was also generally treated as a simple or continuous beam supported by the mode superposition method, and simplified vehicle models were developed without considering the suspension control , , , , , , , , , , . Following the improved simulation of the maglev vehicle and electromagnet interaction, more complicated vehicle/guideway models such as two-dimensional , , , ,  and three-dimensional vehicle models , ,  were developed that consider the proportional-integral-derivative (PID) control  or LQG control theory . These models have been used to investigate more complicated dynamic interactions of maglev systems including the effects of irregularities , , resonant conditions , ground settlement  and cross-wind loads . Field experiments , , , , , ,  were also conducted to investigate the dynamic response of maglev vehicles running on a bridge. The data from the field experiments subsequently enabled the established numerical model to be verified and used to predict the system response. Overall, studies have generally used model-based methods, such as simple analytical or finite element methods, to conduct numerical investigations of the dynamic responses under various conditions, with data from field experiments being used to verify the established models . However, few studies have been conducted on the maglev train-turnout dynamic interactions, as most of the existing research has focused on concrete bridges typically used on main lines , , , . Moreover, field tests conducted on a real operational maglev line are inadequate for turnout.
The turnout is one of the most critical structures in the maglev system to enable the trafficability of the track. To ensure its switching function, the turnout is constructed of steel with very lightweight. However, compared with the normal guideway line, it is a more complex structure. Therefore, the dynamic interaction of the train and the turnout is likely to be more violent than that of train and ordinary concrete bridge, which in turn will further affect the running safety and the riding experience, posing a challenge to the solidity of the turnout. Moreover, it is difficult to characterise the dynamic features of turnout by means of the model-based method. Hence, in this paper, field tests of the Fenghuang maglev sightseeing express line are conducted to study the dynamic performance of the low- and medium-speed maglev train moving on the turnout. This study will provide basic data references for numerical model verification, scientific design, promotion of the related standards.
The reminder of the paper is arranged as follows: Section 2 has a detailed introduction to the information of test train, test line, as well as test scheme; Section 3 discusses the test results of the system’s dynamic responses including vehicle body and the turnout under different operating conditions, and key dynamic index and frequency spectrum characteristics of the system are also analyzed; the relevant conclusions are given in Section 4.
The maglev train (Figure 1) for the field test adopts electromagnetic suspension (EMS) guiding technique and short-stator linear induction motor traction technique. The train has one vehicle body, five suspension bogies and forty electromagnets. Each suspension bogie is decoupled into two side bogies connected by the anti-rolling beams to prevent the suspension bogies from rolling. There are four air springs between the vehicle body and each suspension bogie. The interaction between the F-type rail and the electromagnets is realised through electromagnetic forces, which are controlled by a designated PID controller. The key structural components of the suspension bogies are as shown in Figure 2, and main technical parameters of the maglev train are shown in Table 1.
Train length (mm)
16500 (end), 15600 (middle)
Train width (mm)
Maximum train height (mm)
Net height inside (mm)
Floor height (mm)
AW0 train weight (t)
Train suspension ability (t)
Suspension frame module length (mm)
Rail gauge (mm)
Centerline of coupler’s height (mm)
Number of doors on each side
Starting acceleration (m/s2)
Common braking deceleration (m/s2)
Emergency braking deceleration (m/s2)
Fenghuang maglev sightseeing express line is the first “maglev + culture + tourism” project in the world. Fenghuang maglev project seamlessly connects Zhangjihuai high-speed rail with a total length of 9.121 km. It starts from Fenghuanggucheng Station of Zhangjihuai high-speed railway, passes through the Tourist Service Center in the North of the City, and finally reaches the tunnel entrance of Folklore Park in the direction of 209 Around City Line.
The Fenghuang maglev sightseeing express line includes turnouts, straight sections, small radius curve sections, slop sections and tunnel sections. The function of the turnout is to change the running route of the maglev train from one line to another by the motor-driven long-span girder (LSG). As shown in Figure 3, the turnout of the Fenghuang maglev sightseeing express line is a single joint turnout with three segments i.e., main line, auxiliary line and turnout. There is one short rail joint on the auxiliary line to connect the two F-type rail ends.
To explore the dynamic performance of the test maglev system when the low- and medium-speed maglev train is moving on the turnout, 18 test points are arranged on the vehicle body, suspension bogie, as well as at short rail joint, LSG and continuous beam of the turnout in total. The detailed measurement points for the train and turnout are listed in Table 2.
Furthermore, to fully test the dynamic performance of the train and turnout, different train loads are considered as AW0 (weight of an empty vehicle), AW2 (AW0+weight of seated passengers+weight of standees with 4 people/m2), and AW3 (AW0+weight of seated passengers+weight of standees with 6 people/m2) , and the test speeds are set in a 20 ~ 60 km/h range with an interval of 20 km/h to ensure the safe operation of the test train.
The sampling frequency is set as 5000 Hz, and all the data are collected by an online monitoring system installed on the moving vehicle, track and bridge.
Left side of front end
2# Suspension frame
Left side of front end
Left side of rear end
Right side of front end
Right side of rear end
3# Suspension frame
Left side of front end
Left side of rear end
Right side of front end
Right side of rear end
Short rail joint
Mid-span of the 2nd span
F-rail of the mid-span of the 2nd span
1/4-span of the 2nd span
F-rail of the 1/4-span of the 2nd span
F-rail of the 1st LSG
To investigate the impact of the running speed and the load of the test train on the vibration of vehicle body when the maglev train passes through the turnout, the maximum vertical acceleration amplitudes of the 3 measured points at the vehicle body under different operating conditions are shown in Figure 4. It can be observed that the acceleration amplitudes of the carriage at speed of 20 km/h are the smallest, while the acceleration amplitudes at speed of 40 km/h are a little larger than at other speeds at most cases. The phenomenon is much more obvious with the load condition as AW2. According to reference , it is known that train resonance occurs when a train is moving on a guideway at the resonance speed, it can be concluded that 40 km/h can be the resonance speed for the test system.
It is found that the acceleration of the front and middle points of the vehicle body is larger than that of the middle point due to the pitching motion of the vehicle body. The amplitude of the acceleration is within the range of 0.1772 m/s2 and 0.512 m/s2 at the front point, 0.1131 m/s2 and 0.2698 m/s2 at the middle point, and 0.2429 m/s2 and 0.4387 m/s2 at the rear point of the vehicle body, respectively. To further evaluate the comfort of the vehicle body, Sperling indexes (defined in GB/T 5599-2019) of the three measured points are calculated under different operating conditions. Here, the Sperling index is given by :
where Ai is the vibration acceleration, fi is the vibration frequency, F(fi) is the frequency correlation coefficient given in Table 3.
0.5 ~ 5.4 Hz
5.4 ~ 26 Hz
When the dynamic response contains multiple frequency components, the Sperling index is calculated as:
As shown in Figure 4, it can be seen that the Spering indexes of the maglev train at the location of front and rear are higher than these at the location of middle. All the Spring indexes of the maglev train are lower than 1, which indicates the stability of the maglev train is excellent since all the indexes do not exceed 2.5. Moreover, the train speed and loading have little effect on the comfort of the maglev train.
The dynamic performance of the suspension bogie is also investigated when the maglev train passes through the turnout. The maximum acceleration amplitudes of the measured points installed at 2# and 3# suspension frames under different operating conditions are shown in Figure 5. It is found that the acceleration amplitude of the vehicle body is much smaller than the suspension bogie due to the air spring by comparing the Figure 4 and Figure 5. Meanwhile, as shown in Figure 5, the maximum accelerations of the suspension bogie exhibit similar patterns with the vehicle body, for which the acceleration amplitudes of the running speed as 20 km/h are the smallest basically, and the amplitudes at speed of 40 km/h is larger than the 60 km/h and 20 km/h at most cases with load condition as AW2.
When examing the amplitudes of the suspension frames, it is found out that the amplitudes of the accelerations for the 2# suspension frame is within the range of 1.3943 m/s2 and 2.4205 m/s2 for the left side of front end, 1.2158 m/s2 and 2.9202 m/s2 for the left side of rear end, 0.8212 m/s2 and 3.5616 m/s2 for the right side of front end, and 0.9514 m/s2 and 2.3493 m/s2 for the right side of rear end, respectively. The amplitudes of the acceleration for the 3# suspension frame are within the range of 1.1438 m/s2 and 3.3596 m/s2 for the left side of front end, 1.8099 m/s2 and 2.8168 m/s2 for the left side of rear end, 1.0294 m/s2 and 4.3464 m/s2 for the right side of front end, and 0.7927 m/s2 and 3.0401 m/s2 for the right side of rear end, respectively. The acceleration responses of the 2# and 3# suspension frames are different because they are controlled by different controllers.
To further investigate the dynamic performance of turnout, the maglev train with the load as AW2 moving on the turnout is considered. The maximum acceleration amplitudes of the measured points installed on turnout under different train speeds are shown in Figure 6, and the fast Fourier Transform is performed to analyze the data in the frequency domain and plotted in Figure 7.
As seen in Figure 6, it can be observed that theacceleration amplitudes of the F-rail on the short rail joint and the continuous beam come to be largest when the train speed is up to 40 km/h, which is consistent with the phenomenon observed from Figure 4 and Figure 5. It can be obviously observed from Figure 7 that the resonance occurs at speed of 40 km/h with the resonance frequency being 24.2 Hz. The acceleration of the F-rail on the continuous beam is shown in Figure 8, and the resonance phenomena can also be obviously observed from time domain in the figure. It can also explain the larger acceleration amplitudes of the vehicle at speed of 40 km/h under such operation condition.
In addition, the amplitudes of the acceleration for the short rail joint are within the range of 1.0499 m/s2 and 5.1689 m/s2 for lateral, and between 2.1305 m/s2 and 7.5462 m/s2 for vertical. As for the continuous beam, the amplitudes are within 1.0021 m/s2 and 1.6603 m/s2 for beam, and within 0.6746 m/s2 and 2.1425 m/s2 for the F-rail. For the LSG, they are within 0.2899 m/s2 and 1.0093 m/s2 for the beam, and between 3.0192 m/s2 and 8.2872 m/s2 for the F-rail. It can be concluded that the amplitude of the F-rail is larger than the bridge due to the steel sleepers which function as springs.
Experimental study is conducted on the Fenghuang maglev sightseeing express line, and dynamic performance of the maglev train passing through the turnout section is analyzed in detail in both time and frequency domains. The main conclusions are listed below:
(1) The speed and load of the maglev train affect the acceleration amplitude for both vehicle and turnout section. Generally, the acceleration amplitudes come to be smallest with train speed as 20 km/h, and the amplitudes become largest at train speed of 40 km/h at most cases and especially with train load as AW2, which is caused by the resonance of the turnout;
(2) The vertical Sperling indexes of the test vehicle under all tested operating conditions are less than 2.5 showing that the vehicle is in excellent stability;
(3) Resonance phenomena is more obvious at the F-rail of the short rail joint and continuous beam, and the resonance frequency is 24.2 Hz.
The data supporting our research results are included within the article or supplementary material.
The authors declare no conflict of interest.