0 evaluări0% au considerat acest document util (0 voturi)
65 vizualizări9 pagini
This document analyzes the vertical dynamic wheel/rail interaction caused by polygonal wheels on high-speed trains through a vehicle-track coupled dynamics model. It finds that polygonal wheels result in considerable fluctuation of wheel/rail contact force. The derivative of wheel radial deviations effectively reflects dynamic wheel/rail contact force at speeds above 300 km/h. High order harmonics have a greater influence on dynamic wheel/rail load at higher speeds. Polygonal wheels mainly affect vehicle system through wheelset vibration, and their short wavelengths stimulate high-frequency vibrations in rails and sleepers.
This document analyzes the vertical dynamic wheel/rail interaction caused by polygonal wheels on high-speed trains through a vehicle-track coupled dynamics model. It finds that polygonal wheels result in considerable fluctuation of wheel/rail contact force. The derivative of wheel radial deviations effectively reflects dynamic wheel/rail contact force at speeds above 300 km/h. High order harmonics have a greater influence on dynamic wheel/rail load at higher speeds. Polygonal wheels mainly affect vehicle system through wheelset vibration, and their short wavelengths stimulate high-frequency vibrations in rails and sleepers.
This document analyzes the vertical dynamic wheel/rail interaction caused by polygonal wheels on high-speed trains through a vehicle-track coupled dynamics model. It finds that polygonal wheels result in considerable fluctuation of wheel/rail contact force. The derivative of wheel radial deviations effectively reflects dynamic wheel/rail contact force at speeds above 300 km/h. High order harmonics have a greater influence on dynamic wheel/rail load at higher speeds. Polygonal wheels mainly affect vehicle system through wheelset vibration, and their short wavelengths stimulate high-frequency vibrations in rails and sleepers.
Analysis of vertical dynamic wheel/rail interaction caused by polygonal
wheels on high-speed trains
Xiaoyuan Liu a,b,n , Wanming Zhai a a State Key Laboratory of Traction Power, Southwest Jiaotong University, Chengdu 610031, China b Dipartimento di Meccanica, Politecnico di Milano, Via La Masa 1, Milano 20156, Italy a r t i c l e i n f o Article history: Received 20 November 2013 Accepted 24 November 2013 Available online 1 December 2013 Keywords: Dynamic wheel/rail interaction Polygonal wheel High-speed train Vehicle-track coupled dynamics a b s t r a c t In order to investigate vertical dynamic wheel/rail interaction resulting from a polygonal wheel at high- speeds, a vertical vehicle-track coupled dynamics model is adopted. The measured results from German ICE polygonal wheels are utilized as multi-harmonic excitations. Two types of out-of-round wheels, the periodic and the non-periodic, are simulated and compared. The dynamic wheel/rail interaction induced by different harmonics is analyzed in both time and frequency domains. Results demonstrate that the polygonal wheel will result in considerable uctuation of wheel/rail contact force. Derivative of wheel radial deviations can effectively reect dynamic wheel/rail contact force at train speeds above 300 km/h. Dynamic wheel/rail load becomes more sensitive to high order harmonics with increased speeds. Inuence of the out-of-round wheel on vehicle system is mainly related to the wheelset vibration. Short- wavelength components of the polygonal wheel stimulate high-frequency vibrations of rails and sleepers. & 2013 Elsevier B.V. All rights reserved. 1. Introduction Polygonalization of wheel tread is a common imperfection form among railway vehicle wheels. The wavelength of this defect ranges from 14 cm to the whole circumference, while its ampli- tude is usually less than 1 mm [1]. It is almost impossible to detect this defect by visual inspection. However, the polygonal wheel poses a potential threat to operational safety, especially for high- speed trains. In 1998, an ICE high-speed train derailed and crashed to a bridge near Eschede, resulting in devastating consequences. The investigation results revealed that strong wheel/rail contact loads due to a polygonal rubber-spring wheel broke the wheel rim, leading to derailment. In order to ensure the operating safety of high-speed trains and reduce the damage to both vehicles and tracks, it is essential to execute timely inspection, re-proling and removal of wheels. Many valuable researches were conducted in this eld. Wheel polygonalization resulted from wheelset unbalance and rotational inertia appeared in [2,3]. To simulate development of polygonal wheels, the coupling of a vehicle-track dynamics model and a wheel wear model by iteration in different time scales was accomplished in Refs. [47]. Measurements of wheel tread irregularity and dynamic tests of wheel tread defects were carried out in [8,9]. Refs. [10,11] introduce wheelset elastic resonance and wheel/rail load problems due to polygonalized wheels. In previous studies on the vehicle/track dynamic interaction due to wheel irregularity, mono-harmonic is widely adopted as the excitation form among many papers. Nevertheless, the actual out- of-round wheel has very complicated geometry. And the dynamic wheel/rail contact force does not have a direct connection with harmonic orders or vehicle speeds. In contrast, the excitation form of multi-harmonic is more effective for analyzing wheel/rail interaction induced by the realistic polygonal wheel. This paper is inclined to present the inunce of wheel out-of- roundness on dynamic behaviours of vehicle-track system by considering the characteristics in actual polygonal wheels of high-speed trains. 2. Simulation model 2.1. Vehicle-track coupled dynamics model On basis of the framework of vehicle-track coupled dynamics [12], a vehicle-track vertical model is built to analyze the dynamic wheel/rail interaction due to polygonal wheels at high-speed operations. In the model, the vehicle system is regarded as a multi-body dynamic system. The vertical and pitch motion of carbody and two bogie frames and the vertical motion of four Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/wear Wear 0043-1648/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.wear.2013.11.048 n Corresponding author. E-mail address: xiaoyuan.liu@polimi.it (X. Liu). Wear 314 (2014) 282290 wheelsets are taken into consideration. The track system is modeled as an Euler beam exibly supported by a discrete-elastic foundation of three layers with sleepers and ballasts included. The ballast bed is separated into blocks corresponding to the sleeper spacing. The coupling of the vehicle system and the track system is achieved by the vertical wheel/rail contact force. The simplied Hertz non-linear elastic contact model is utilized to calculate the wheel/rail contact force. A simple fast time integration algorithm constructed by Zhai [13] is employed to solve the high dimensional differential equations of the vehicle-track system. 2.2. Mathematical model of polygonal wheel For the purpose of analyzing the vertical dynamic wheel/rail interaction caused by the out-of-round wheel, one solution is to convert circumferential irregularity into harmonic excitation of the wheel/rail interface. The wheel radial deviation Z w0 () can be expressed as Fourier series with N harmonics: Z w0
N i 1 A i sini i 1 where is a polar coordinate, i is order of the polygon, A i is corresponding amplitude, and i is a phase. Figs. 1 and 2 display polar coordinate graphs and harmonic distributions of two types of polygonal wheels from [1]. Polar coordinate diagrams are tted with the method mentioned above, based on the detected data from German ICE1 wheels. Wheel I is a periodic polygonal wheel whose dominant harmonics are the 1st and 3rd orders. Fixation misalignment in manufacture and assembly may result in the common 1st order, while clamping in a three-jaw chuck during proling may lead to the 3rd order. Wheel II is a non-periodic (stochastic) polygonal wheel without denite dominant orders, but it has more short-wavelength components. Its formation might be related to heat treatment [14]. The radial deviation Z w0 () can be converted into the time history: Z w0 t N i 1 A i sin iV=Rt i 2 where V represents the vehicle speed; R represents the mean radius. According to Eq. (2), the rst derivative of Z w0 (t) with respect to t can be obtained: dZ w0 t dt V=R N i 1 iA i cos iV=Rt i 3 Without considering track irregularities, the vertical compres- sion amount between wheel and rail can be written as Zt Z w tZ w0 tZ r x w ; t 4 where Z w (t) is the wheel vertical displacement, Z r (x w , t) is the displacement of the rail beneath the wheel. The vertical wheel/rail contact force can be determined by use of the simplied Hertzain non-linear contact theory: Pt Zt=G 3=2 ZtZ0 0 Zto0 ( 5 where G is the wheel/rail contact constant related to the wheel radius and tread type. 3. Results and analyses Vehicle dynamics parameters of a Chinese high-speed EMU are used in the simulation. Ballasted track parameters from a passenger-dedicated line are adopted correspondingly. The main dynamics parameters are listed in Table 1. Irregularity of the Fig. 1. Wheel I (peak-to-peak value 0.44 mm): (a) polar diagram and (b) harmonic distribution. Fig. 2. Wheel II (peak-to-peak value 0.35 mm): (a) polar diagram and (b) harmonic distribution. X. Liu, W. Zhai / Wear 314 (2014) 282290 283 polygonal wheel is imposed on the steering wheelset of the leading bogie. Based on the polygonal geometry illustrated above, dynamic responses are analyzed under different operation speeds. 3.1. Dynamic wheel/rail contact force Fig. 3 demonstrates responses of wheel/rail contact force for polygonal wheels at train speed of 350 km/h. It is noticed from Fig. 3 that the wheel/rail force shows considerable periodic uctuations with radial deviations. Peaks of the contact force occur at the raising slopes of radius, while troughs happen at falling slopes. Actually, the damping effect of the vehicle-track coupled system takes an important role in dynamic wheel/rail contact force. Due to the high stiffness of wheel/rail contact, the rapid variations of wheel irregularities induce signicant oscilla- tions of the wheel/rail compression and result in strong dynamic contact loads. Fig. 4 displays frequency spectra of the wheel radial deviation, dynamic wheel/rail displacement, and dynamic contact force for wheel II at 150 km/h. The dynamic wheel/rail interaction can be classied into three categories according to excitation frequencies. For the 1st order, the wheel bounces with its own irregularity, leading to very low contact load. It corresponds to frequencies below 20 Hz. As to the 2nd to 4th orders, both wheel and rail show evident vibration, resulting in signicant wheel/rail loads. This phenomenon is caused by wheel/track coupled vibration ranging from 20 Hz to approximate 120 Hz. With regard to higher order harmonics, these components cause elastic deformation of the rail; however, wheel displacement amplitude keeps low. They also generate evident contact loads. It is mainly due to high-frequency wheel/rail Hertz contact vibration above 120 Hz. In addition, the dynamic rail displacement reveals intensity of the wheel/rail contact load in frequency domain. As the vehicle speed increases, the wheel/track coupled vibration and wheel/rail high-frequency Hertz contact vibration become more dominating. Especially the majority of the harmonics induces the wheel/rail Hertz contact vibration when the vehicle speed exceeds 300 km/h. Conse- quently, the wheel/rail contact force at high speed is closely associated with the radial geometry of the wheel. Table 1 Main dynamics parameters used in simulation. System Parameter Value Vehicle Axle load (kg) 11200 Unspring mass (kg) 2100 Wheel rolling radius (m) 0.43 Track Rail mass (kg/m) 60 Sleeper spacing (m) 0.6 Rail pad stiffness (N/m) 6.010 7 Support stiffness of ballast bed (N/m) 2.210 8 Fig. 3. Time history of wheel/rail contact force and wheel radial deviation at 350 km/h: (a) Wheel I and (b) Wheel II. X. Liu, W. Zhai / Wear 314 (2014) 282290 284 Since the wheel/rail contact force is sensitive to the variation of wheel radial deviation, derivative of radial deviation with respect to time (could also be length or angle) is a proper indicator of the dynamic load. It means the derivative of radial deviation is regarded as the velocity corresponding to the damping of the vehicle-track system. Fig. 5 displays time history of the wheel/rail force and the rst derivative of radial deviations at 350 km/h. In Fig. 5, the derivative is almost consistent with the wheel/rail contact force. The peaks are located at maxima of the derivative, and the trough values stay at the minima. Fig. 6 shows the frequency spectra of the wheel/rail load and derivative of radial deviation. The contact force and derivative of deviation also display similar patterns among most orders. Figs. 7 and 8 exhibit additional dynamic load and wheel load reduction ratio versus speed, respectively. The wheel load reduc- tion ratio, dened as (P static P min )/P static , is a safety index to evaluate the severity of the wheel unloading, which may induce a jumping derailment. Generally, as the train speed grows, uctuation of wheel/rail contact force shows a noticeable increased trend. Nevertheless, it does not mean that the dynamic force due to polygonal wheels monotonously intensies with higher speeds. A local maximum can be found at 150 km/h for Wheel I, probably because the dominant 3rd order induces P2 resonance around 45 Hz [6,15,16]. At this frequency, wheel and rail almost vibrate in phase. In frequency domain, Fig. 9 indicates each harmonic's contribu- tion to the dynamic wheel/rail force at different speeds. As for Wheel I, the dominant 3rd order harmonic can lead to signicant dynamic load responses in a wide range of speeds over 150 km/h. It is necessary to make the 3rd order under control in the process of proling and re-proling. For Wheel II, the dynamic loads become more sensitive to both higher and lower order harmonics, as the speed increases. Higher orders intensify Hertz contact with incremental frequencies; lower orders of high amplitude gradually induce wheel/track coupled vibration with the increased speeds. There is a convex zone for 8th10th orders over 300 km/h, however, the current criteria (peak-to-peak value) underestimate hazards of high order harmonics, whose amplitudes are always low. It also appears that the conditions of 300400 km/h display similar patterning due to the wheel/rail Hertz contact. Therefore, Fig. 4. Spectra of wheel radial deviation, dynamic wheel/rail displacement and contact load for Wheel II at 150 km/h. X. Liu, W. Zhai / Wear 314 (2014) 282290 285 the derivatives of wheel radial deviation can demonstrate dynamic wheel/rail contact loads at high speeds, particularly for the short- wavelength harmonics. 3.2. Acceleration response of vehicle system Wheelset acceleration resulting from the out-of-round wheel reects characteristic of the dynamic wheel/rail contact load, as illustrated in Fig. 10. A probable cause is that dynamic component of the primary suspension force is substantially weaker than that of the wheel/rail contact force at high frequencies. Fig. 11 displays the trend of wheelset acceleration with train speeds, exhibiting consistent tendency compared with Fig. 7. The high frequency oscillation may shorten the lifespan of axlebox bearings, hydraulic dampers, or even cause fatigue of bogie frame. Figs. 12 and 13 respectively show accelerations of the carbody and the leading bogie frame at various train speeds. The acceleration responses of both carbody and frame stay at low levels at all speeds. Owing to the low natural frequencies and damping ratios of primary and secondary suspensions, high frequency oscillations from the wheel can be effectively attenuated. However, the structural vibra- tion of the carbody may be amplied by polygonal wheels, which is already illustrated in several Refs. [5,10,11]. 3.3. Acceleration response of track system Fig. 14 exhibits time history and frequency spectra of rail acceleration at 350 km/h. In contrast with wheel acceleration, the rail acceleration contains distinct high frequency components. Response of Wheel II is far more intense than Wheel I, because short-wavelength harmonics of Wheel II excite the high-order eigenmodes of rail. Fig. 15 demonstrates the maxima of the rail acceleration with speeds. There is a denite discrepancy between the two wheels: as train speed exceeds 250 km/h, the rail acceleration for Wheel II shows a noticeable increased tendency compared with Wheel I. The stochastic polygonal wheel induces more violent rail vibration than the periodic polygonal wheel, even though their peak-to- peak values are approximate. Fig. 16 illustrates the contributions of the harmonics to the rail acceleration. Rail vibration resulting from the high-order harmonics show a drastic enlargement with raising speeds, however, vibration from the low-order harmonics remains relatively low. When the train speed exceeds 300 km/h, the excitation frequencies from high-order harmonics come close to the eigenfrequency of rail bouncing over rail pads [17], leading to apparent high-frequency responses. Therefore, the subtle high-order harmonics play a critical role in rail vibration. Fig. 17 exhibits the acceleration of sleeper and ballast bed at different speeds. In contrast to Wheel I, the sleeper acceleration Fig. 5. Time history of wheel/rail contact force and derivative of wheel radial deviation at 350 km/h: (a) Wheel I and (b) Wheel II. X. Liu, W. Zhai / Wear 314 (2014) 282290 286 for Wheel II also shows a steeper gradient over 250 km/h, mainly due to high-frequency rail oscillation. As the speed exceeds 300 km/h, the acceleration of ballast bed remains comparably stable. Fig. 6. Frequency spectra of wheel/rail contact force and derivative of radial deviation at 350 km/h: (a) Wheel I and (b) Wheel II. Fig. 7. Additional dynamic wheel/rail load due to polygonal wheels versus train speed. Fig. 8. Wheel load reduction ratio due to polygonal wheels versus train speed. Fig. 9. Dynamic wheel/rail contact force caused by each harmonic at various train speeds: (a) Wheel I and (b) Wheel II. X. Liu, W. Zhai / Wear 314 (2014) 282290 287 4. Conclusions A high-frequency vehicle and track coupled dynamics model is employed to simulate dynamic wheel/rail interaction due to polygonal wheels at high speeds. The multi-harmonic excitation is adopted to reect the irregularities of actual polygonal wheels. Based on this model, the vehicle-track interactions induced by the periodic and stochastic polygonal wheels are compared in time and frequency domains. Fig. 10. Response of wheelset acceleration at 350 km/h: (a) Wheel I and (b) Wheel II. Fig. 11. Maxima of wheelset acceleration versus train speeds. Fig. 12. Peak values of leading bogie frame acceleration. Fig. 13. Peak values of carbody acceleration. Fig. 14. The response of rail acceleration: (a) Wheel I and (b) Wheel II. X. Liu, W. Zhai / Wear 314 (2014) 282290 288 Simulation results demonstrate that the polygonal wheel will result in substantial uctuations of vertical wheel/rail contact force. The dominant 3rd order harmonic leads to signicant dynamic load in a wide range of speeds, so the initial 3rd order due to proling or re-proling should be kept under reasonable control. For non-periodic polygonal wheels, the dynamic wheel/ rail contact force is more sensitive to both higher and lower order harmonics at higher speeds. The wheel/rail dynamic interaction can be categorized into three forms according to the excitation frequencies. As the wheel/rail Hertz contact vibration is dominant at high speeds, derivative of wheel radial deviations is an efcient indicator of dynamic wheel/rail contact force. It is recommended that derivatives of wheel radial deviations should be incorporated in the criteria for re-proling and removal. For the vehicle system, polygonal wheels mainly affect the wheelset vibration. The vehicle suspensions can effectively attenu- ate the resultant vibration. The short-wavelength harmonics of the polygonal wheel considerably intensify high frequency oscillations of rails and sleepers, especially at higher speeds. Acknowledgements This research was supported by the National Natural Science Foundation of China (NSFC) under grants U1234209 and by National Key Basic Research Program of China (973 Program) under grant 2013CB036206. References [1] J. Nielsen, A. Johansson, Out-of-round railway wheelsa literature survey, Proc. Inst. Mech. Eng. Part F J. Rail Rapid Transit 214 (2000) 7991. [2] E. Brommundt, A simple mechanism for the polygonalization of railway wheels by wear, Mech. Res. Commun. 24 (1997) 435442. [3] P. Meinke, S. Meinker, Polygonalization of wheel treads caused by static and dynamic imbalances, J. Sound Vib. 227 (5) (1999) 979986. [4] B. Morys, Enlargement of out-of-round wheel proles on high-speed trains, J. Sound Vib. 227 (5) (1999) 965978. [5] T. Meinders, P. Meinke, W. Schiehlen, Wear estimation in exible multibody systems with application to railroads, in: Multibody Dynamics, Eccomas Thematic Conference, Madrid, 2005. [6] A. Johansson, C. Andersson, Out-of-round railway wheelsa study of wheel polygonalization through simulation of three-dimensional wheelrail interac- tion and wear, Veh. Syst. Dyn. 43 (8) (2005) 539559. [7] M. Meywerk, Polygonalization of railway wheels, Arch. Appl. Mech. 69 (1999) 105120. [8] A. Johansson, Out-of-round railway wheelsassessment of wheel tread irregularities in train trafc, J. Sound Vib. 293 (2006) 795806. [9] A. Johansson, J. Nielsen, Out-of-round railway wheelswheelrail contact forces and track and track response derived from eld tests and numerical simulations, Proc. Inst. Mech. Eng. Part F J. Rail Rapid Transit 217 (2003) 135146. Fig. 15. Maxima of rail acceleration versus train speed. Fig. 16. Rail acceleration caused by each harmonic at different train speeds: (a) Wheel I and (b) Wheel II. Fig. 17. Maxima of sleeper and ballast bed accelerations. X. Liu, W. Zhai / Wear 314 (2014) 282290 289 [10] K. Popp, I. Kaiser, H. Kruse, System dynamics of railway vehicles and track, Arch. Appl. Mech. 72 (2003) 949961. [11] H. Claus, W. Schiehlen, H. Kruse, Dynamic stability and random vibrations of rigid and elastic wheelsets, Nonlinear Dyn. 36 (2004) 299311. [12] W.M. Zhai, K.Y. Wang, C.B. Cai, Fundamentals of vehicle-track coupled dynamics, Veh. Syst. Dyn. 47 (11) (2009) 13491376. [13] W.M. Zhai, Two simple fast integration methods for large-scale dynamic problems in engineering, Int. J. Numer. Methods Eng. 39 (24) (1996) 41994214. [14] J. Nielsen, R. Lundn, A. Johansson, T. Vernersson, Train-track interaction and mechanisms of irregular wear on wheel and rail surfaces, Veh. Syst. Dyn. 40 (2003) 354. [15] S. Bruni, F. Braghin, Effect of damage on vehicle dynamics, in: R. Lewis, U. Olofsson (Eds.), WheelRail Interface Handbook, Woodhead Publishing Limited and CRC Press LLC, Cambridge, 2009, pp. 456474. [16] J. Nielsen, Out-of-round railway wheels, in: R. Lewis, U. Olofsson (Eds.), WheelRail Interface Handbook, Woodhead Publishing Limited and CRC Press LLC, Cambridge, 2009, pp. 245275. [17] S. Bruni, J. Vinolas, M. Berg, O. Polach, S. Stichel, Modeling of suspension components in a rail vehicle dynamics context, Veh. Syst. Dyn. 49 (7) (2011) 10211072. X. Liu, W. Zhai / Wear 314 (2014) 282290 290