1、在风雨情况下的高速列车空气动力学建模和稳定性分析 英文版Shao et al. / J Zhejiang Univ-Sci A (Appl Phys & Eng) 2011 12(12):964-970964Aerodynamic modeling and stability analysis of a high-speed trainunder strong rain and crosswind conditions*Xue-ming SHAO1, Jun WAN1, Da-wei CHEN2, Hong-bing XIONG1(1Department of Mechanics, Zheji
2、ang University, Hangzhou 310027, China)(2National Engineering Laboratory for System Integration of High-Speed Train (South),CSR Qingdao Sifang Co., Ltd., Qingdao 266111, China)E-mail: hbxiongReceived Sept 23, 2011; Revision accepted Sept 24, 2011; Crosschecked Sept 24, 2011Abstract: With the develop
3、ment of high-speed train, it is considerably concerned about the aerodynamic characteristics andoperation safety issues of the high-speed train under extreme weather conditions. The aerodynamic performance of a high-speedtrain under heavy rain and strong crosswind conditions are modeled using the Eu
4、lerian two-phase model in this paper. The impact of heavy rainfall on train aerodynamics is investigated, coupling heavy rain and a strong crosswind. Results show that the lift force,side force, and rolling moment of the train increase significantly with wind speed up to 40 m/s under a rainfall rate
5、 of 60 mm/h.when considering the rain and wind conditions. The increases of the lift force, side force, and rolling moment may deteriorate thetrain operating safety and cause the train to overturn. A quasi-static stability analysis based on the moment balance is used todetermine the limit safety spe
6、ed of a train under different rain and wind levels. The results can provide a frame of reference for thetrain safe operation under strong rain and crosswind conditions.Key words: High-speed train, Aerodynamic characteristics, Multiphase flow, Rain, Crosswind, Overturningdoi:10.1631/jzus.A11GT001 Doc
7、ument code: A CLC number: O359; TD5241 IntroductionWhen high-speed trains running under strongrain and crosswind conditions, especially at exposedlocations such as bridges or embankments, theaerodynamic forces and moments may increasesignificantly and result in the train instability. It was well-kno
8、wn that the strong crosswind may increasethe aerodynamic drag force, side force and yawingmoment. If the rain and the crosswind coexist, theaerodynamic performance of train will deterioratemore severely, which may cause train delays,shutdowns, derailments and even overturning. In2007, trains deraile
9、d with 11 cars under high windconditions, causing a serious accident on the XinjiangRailway South Line and a train overturned in Tianjin(Fig. 1a). In Japan, there have been about 30wind-induced accidents to date (Fig. 1b).Numerous studies have been done on the aero-dynamic performance of trains unde
10、r crosswind con-ditions. Ma et al. (2009) investigated the aerodynamiccharacteristics of a train on a straight line at 350 km/hwith a crosswind. More systematic researches onnumerical simulation of a train travelling along astraight line and also curves have been done (Liangand Shen, 2007; Yang et a
11、l., 2010). In addition towind tunnel experiments, numerical simulation ana-lyzes the effects of crosswinds in more detail withresults consistent with those from experiments(Christina et al., 2004; Javier et al., 2009; Sanquer etal., 2004; Masson et al., 2009). However, less atten-tion has been paid
12、to the effects of combined strongrain and crosswind on the aerodynamic characteristicsJournal of Zhejiang University-SCIENCE A (Applied Physics & Engineering)ISSN 1673-565X (Print); ISSN 1862-1775 (Online) E-mail: jzus Corresponding author* Project (No. 2009BAG12A01-C03) supported by the National Ke
13、yTechnology R&D Program of China Zhejiang University and Springer-Verlag Berlin Heidelberg 2011Shao et al. / J Zhejiang Univ-Sci A (Appl Phys & Eng) 2011 12(12):964-970965and safety of high-speed trains. This paper deals withthe influence of strong rain and crosswind conditionson high-speed train ae
14、rodynamics, based onnumerical simulations. A quasi-static stability analy-sis based on the moment balance is also used to de-termine the limit safety speed of a train under dif-ferent rain and wind levels, which provides someguidance for the train operation safety.2 Numerical simulation2.1 Computati
15、onal modelComputational fluid dynamics (CFD) softwareFLUENT is used for numerical simulation in thisstudy. For multiphase flow problems, there are mainlytwo types of multiphase flow models: one is thediscrete particle model (DPM) proposed by Crowe andSmoot (1979); the other one is Eulerian-Eulerianm
16、odel proposed by Gidaspow (1994). The fluid phasein DPM is solved using the Eulerian method, while thegranular phase is tracked by the Lagrangian method.Though providing more detail, DPM is not suitable forlarge-scale engineering simulation with numerousdispersed particles, due to the limitation of
17、finitememory capacity and CPU efficiency. The Eulerian-Eulerian approach is more efficient and usually morecomplex. Each phase is treated as a continuous me-dium that may interpenetrate with other phases, and isdescribed by a set of equations with regard to mo-mentum, continuity, and energy. The Eul
18、erian-Eulerian approach has been successfully applied tothe simulation of gas-particle multiphase flow with alarge number of particles in large equipment. For example, Liu et al. (2006) studied the liquid-solidslurry transport within a pipeline, and Cao et al.(2005) simulated the bubble growth, inte
19、gration, andthe flow characteristics in a fluidized bed. Due tothese advantages, the Eulerian-Eulerian model is usedto simulate the example in this study.The flow around the train is viscous, turbulent,and gas-droplet two-phase flow. Thus, the Eulerian-Eulerian multiphase model coupling with the k-e
20、turbulence equation is used for the gas-droplettwo-phase flow field around the train. Details of theEulerian-Eulerian multiphase model and its validationcould be found in another paper of this issue (Xiong etal., 2011) conducted in our group.2.2 Computational domainA simplified model of CRH2 (China
21、RailwayHigh Speed 2) is studied, including three coaches ofthe head, middle and tail. To fully develop flowaround the train and to ensure the accuracy of results,a large semi-cylindrical numerical wind tunnel wasestablished as the computational domain (Fig. 2). Thedistances of the semi-cylindrical c
22、omputational do-main in the vertical, horizontal and vertical directionsare 600, 400 and 200 m, respectively. The length ofvehicle is 76 m. The distance between nose of vehicleand the inlet boundary is 100 m. Direction of the flowfield is at the positive x-axis.2.3 Computational meshThe computationa
23、l domain is meshed with thehexahedral structured grid, with refine mesh on the(a)Fig. 1 Train overturning in China (a) and Japan (b)(b)Fig. 2 Computational domainABCDxyz400 m600 mFEv200 mOShao et al. / J Zhejiang Univ-Sci A (Appl Phys & Eng) 2011 12(12):964-970966front and rear of the train and surr
24、ounding areas asshown in Fig. 3. To validate the train model vianumerical simulation, grid dependency has beenconducted with three kinds of grid generation:330 000, 600 000, 980 000 grids, as shown in Table 1,where the change rate is defined as,iFFFD-=(1)where Fi is the aerodynamic force (or moment)
25、 fordifferent mesh models, and F is the aerodynamicforce with 98 000 grids. The results of 980 000 gridsare very close to that of 60 000 grids. Therefore, wecan conclude that 980 000 grids are acceptable for thenumerical simulation of the train.2.4 Boundary conditionsPlanes DEF and ABCFED in the com
26、putationaldomain (Fig. 2) are given as the boundaries ofvelocity inlet. Plane ABC is a pressure outletboundary with a static pressure of 0. Train surfacesare stationary, with non-slip boundary condition.Plane ACFD adopts a moving boundary withspeed equal to flow velocity. The crosswind directionis o
27、n the positive y-axis, perpendicular to the train.The continuous phase and the granular phase sets areas follows: the continuous phase uses inlet boundaryvelocity of 360 km/h; the granular phase uses inletboundary with x-axis velocity equal to the gas inletvelocity and negative z-axis velocity of ra
28、indrop of5 m/s. Turbulent kinetic energy k and turbulentdissipation rate are determined by2m0.004ku=(2)1.50.090.03kRe =, (3)where um is the mean flow velocity, and R is theturbulence length scale. The Phase-Coupled-Simplealgorithm is used for solving the coupling betweenpressure and velocity effects
29、 (Moukalled et al., 2003).3 Problem descriptionIn this study, the train running speed is set as360 km/h. The crosswind speed ranges from 0 to40 m/s, and the direction of wind is perpendicular tothe running direction of train. Under the crosswindand heavy rainfall computing conditions, rainfall rate(
30、rainfall intensity per hour) is 60 mm/h. A raindrop isregarded as spherical with constant falling velocity of5 m/s and diameter of 0.002 m.4 Results and discussion4.1 Pressure distribution on the train surfaceThe drag force on the train consists of pressuredrag and friction drag. The pressure drag i
31、ncreasesdramatically when running at high speed. Due to theimpact of rainfall, the pressure distribution around theTable 1 Aerodynamic comparison in different grid modelsMesh modelDrag force (N)Side force (N)Lift force (N)330 00019945.187.6%56.9660.3%7686.271.2%600 00019018.512.6%36.272.1%7747.460.4%980 00018528.52035.5307782.270%Mesh modelRolling moment(Nm)Pitching moment(Nm)Yawing moment(Nm)33
