Home    Reference Manual    Return   



Simulation of a railway vehicle passing a switch


Author: Ingemar Persson (DEsolver)

blurulr2.gif


Introduction

blurulr2.gif

To maintain switches costs a lot of money. In Sweden the maintenance cost for switches stands for a big part of the total track maintenance costs. Because of that, there is of great interest to better understand the mechanisms that destroys the switch. This presentation shows results from a mathematical model of switch and vehicle, when the vehicle passes over at constant speed.

blurulr2.gif


The switch

The switch geometry used in input data has been created by Elias Kassa Chalmers, where he is using the model in his work with his Doctoral Thesis. The curve radius of the switch is 760 [m], without transition curves. The crossing angle is 1:15. Considerations to varying rail profiles has been made. The geometry of all rail profiles are taken from a new unworn switch. In total the switch consists of 31 sections, 14 sections describes the blade, and 17 sections describes the crossing. No track irregularities has been considered.

blurulr2.gif


The vehicle

The vehicle used is taken from the the railway vehicle dynamic course at KTH, conducted by Prof. Mats Berg. The vehicle is a four-axle bogie vehicle, with a axle load of 20 metric tonnes. The speed of the vehicle was 70 [km/h].

blurulr2.gif


Results

Overview, showing the vehicle turning right in a right handed switch.
Start animation

Average contact stress when the when the wheels are hitting the blade of the switch.
The contact stresses in the contact points between wheel and rail are very high. Results from the switch simulation shows that we often have contact stresses above max yield stress limit, this will lead to plastic deformations in the contact surfaces, and the contact stress distribution will not be Hertzian. Therefore have the contact stresses in the animation been presented as the average contact stress over the whole contact surface. In the cases where Hertzian contact stress distribution do exist, the reader can manually increase the contact stress by 50%, in order to obtain maximum contact stress according to Hertz.
The average contact stress on tangent track for this vehicle is about 990 [MPa]. The average contact stress for leading outer wheel in a 760m curve for this vehicle is about 1700 [MPa]. Maximum average contact stress when the leading outer wheel hits the blade is 4760 [MPa] which can be seen in frame #148. The high contact stress that occurs when the leading outer wheel hits the blade will lead to headchecks on the blade, however the wear is also very high which will lead to a removal of all cracks.
Start animation

Energy dissipation when the when the wheels are hitting the blade of the switch.
The energy dissipation on tangent track for this vehicle is about 1 [J/m]. The energy dissipation for leading outer wheel in a 760m curve for this vehicle is about 150 [J/m]. Maximum energy dissipation when leading outer wheel hits the blade is 1072 [J/m], which can be seen in frame #190.
Start animation

Average contact stress when the when the wheels are hitting the crossing nose.
In the animation it can be seen that the wing rail goes down when the crossing nose comes up, because of that the wheel will have a vertical speed when it hits the crossing nose. Despite of the low vertical speed of the wheel when it hits the nose, the dynamic vertical track force will be approx. 50 % of nominal static load. The reason for this high dynamic vertical track force is because of the stiff design of the switch. In addition to the high contact force the lateral curvature difference between wheel and nose is also very high, which leads to a small contact surface. These two effects gives rize to a very high contact pressure when the wheel hits the crossing nose. In frame #1224 it can be seen that the average contact stress can be as high as 6919 [MPa]. The high contact stress can lead to plastic deformation and spalling of the crossing nose.
Start animation

blurulr2.gif


Conclusions

The simulation shows big contact forces on both switch-blade and crossing nose. According to the simulations it appears that the blade of the switch could have a more advantageous geometry, in order to reduce wear and contact stresses on the blade. Regarding the crossing nose maybe it is more difficult to find a better geometry, the vertical kink may be necessary in order to manage many different wheel profiles. Only this 1:15 R=760 switch with vertical standing rails has been analyzed, a switch with inclined rails can have a better geometry.

blurulr2.gif