ANALYTICAL AND NUMERICAL ANALYSIS OF HIGH-SPEED TRAIN MOVES IN

TUNNEL

ABSTRACT

The study examines the aerodynamics of fast trains in tunnels. It uses ANSYS to
simulate compressible, inviscid air flow with a Mach value of 1.7 around the train while
accounting for differences in density. Boundary conditions for temperature and
pressure are established at the intake, outflow, and walls, and energy calculations are
supplied. 5000 simulation iterations and a residual of 0.0004 are required for
convergence. This method validates CFD for compressible flow analysis in high-speed
rail applications by closely matching analytical predictions.

Table of Contents

1. INTRODUCTION AND REVIEW OF LITERATURE ................................................................ 2

2. METHOD ..................................................................................................................................... 2
2.1 Analytical method ................................................................................................................. 2
2.2 Numerical method ................................................................................................................ 4
3. RESULTS .................................................................................................................................... 6
4. CONCLUSION ............................................................................................................................ 8
5. REFERENCES ........................................................................................................................... 8

List of Tables
Table 1 Speed of train using compressible and incompressible flow conditions ....................... 6
Table 2 Comparison of analytical results and CFD result ........................................................... 7

List of Figures
Figure 1 CFD model ....................................................................................................................... 4
Figure 2 Meshing ........................................................................................................................... 5
Figure 3 Pressure contours ........................................................................................................... 7
Figure 4 Velocity contours.............................................................................................................. 7

1
1. INTRODUCTION AND REVIEW OF LITERATURE

The operation of high-speed trains within tunnels presents a critical area of

investigation owing to the complex dynamics arising from the interaction between the
train and the tunnel environment [1]. With the burgeoning development of high-speed
rail networks globally, understanding the intricate dynamics of these systems becomes
imperative for ensuring safety, efficiency and optimal performance [2]. The study aims
to conduct an exhaustive analysis, encompassing both analytical and numerical
approaches, to comprehensively evaluate the dynamics of high-speed trains
traversing through tunnels. The complex interplay of aerodynamics, structural
mechanics and fluid dynamics within the confined space of a tunnel poses unique
challenges that demand a multifaceted analysis [3]. Analytical methodologies will draw
upon established principles of fluid dynamics [4], structural mechanics [5] and train
dynamics equations [6] providing a theoretical framework for understanding the
fundamental behaviour of trains in tunnels. Concurrently, numerical simulations
employing computational fluid dynamics (CFD) [7] and finite element analysis (FEA)
[8] will enable a detailed investigation into the aerodynamic interactions and structural
responses under various operating conditions.

2. METHOD

2.1 Analytical method

First, the analytical equation to calculate the speed of the train under compressible
and incompressible flow assumptions is developed.

Assumptions:

We have a levitating carriage (train) inside a circular tube. The height of the carriage
(C) is fixed, the gap (h) between the carriage roof and tube ceiling varies, 0.05 < h*<
0.50, where h*=h/C. The fluid inside the tube is air and the train travels at a high speed
of about 300 km/hr. The non-dimensionalization such as Length is scaled by the tube
diameter (D), Velocities is scaled by the characteristic sound speed (c0) in the
undisturbed air, Pressure is scaled by the reference pressure (p0) outside the tube
and Density is scaled by the reference density (ρ0) outside the tube.

2
The Bernoulli equation [9] is one of the fundamental laws of fluid dynamics giving rise
to the relationship between pressure total pressure and static pressure. This theorem
states that for an incompressible inviscid flow along a streamline, the total mechanical
energy i.e., kinetic energy + pressure remains constant. This can be applied to two
points along the flow around the carriage;

 Point-1: Upstream of the carriage, where the flow is undisturbed (velocity= U∞,
Pressure p∞).
 Point-2: At the narrowest section of the flow near the carriage roof, where the
velocity increases (V2) due to the reduced flow area and pressure drop (p 2)

Applying Bernoulli’s theorem:

1/2ρ0(U∞^2) + p∞ = 1/2ρ0(V2^2) + p2 Equation 1

Since the flow is assumed incompressible (Ma < 0.3), ρ0 remains constant. We can
solve for V2:

V2 = U∞ * sqrt((1 + (2p∞ / ρ0U∞^2) * (h*/ (1 - h*)))) Equation 2

By analysing the equation for V2 and varying the gap ratio (h*), an increase in velocity
and pressure drop can be calculated.

In defining reference values, it is crucial to select the appropriate values for p0, ρ0,
and U∞ aligned with the particular train speed and air properties under consideration.
Once these reference values are established, non-dimensional parameters like Ma
(Mach number) and Re (Reynolds number) are determined for the chosen train speed
and tube diameter. Use the equation for V2 and Bernoulli’s theorem to analyse how
the flow velocity and pressure drop near the carriage roof vary with different gap ratios
(h*) providing insight into the aerodynamic behaviour and pressure distribution in the
tunnel as the train traverses through it.

This method is valid only for Ma less than 0.3. Compressibility effects become
significant at higher speeds and Bernoulli’s theorem may not be accurate. We can
evaluate the alternate methods such as CFD simulations for more accurate
compressible flow analysis.

The relationship between the turn angle (𝛿) and the wave angle (𝛽) is given by;

3
 2 𝑐𝑜𝑠 𝛽 (𝑀12 𝑠𝑖𝑛 2𝛽−1)
𝑡𝑎𝑛 𝛿 = Equation 3
2+𝑀12 (𝛾+𝑐𝑜𝑠 2𝛽)

Pressure ratio across the oblique shock wave is,

𝑃2 2𝛾𝑀12 𝑠𝑖𝑛 2 𝛽−(𝛾−1)

= Equation 4
𝑃1 (𝛾+1)

Density ratio across the oblique shock wave is;

𝜌2 (𝛾−1)𝑀 2𝑠𝑖𝑛 2 𝛽
= 2+(𝛾−1)𝑀1 2𝑠𝑖𝑛2 𝛽 Equation 5
𝜌1 1

Temperature ratio across the oblique shock wave is

𝑇2 (2+(𝛾−1)𝑀12𝑠𝑖𝑛 2 𝛽)(2𝛾𝑀12 𝑠𝑖𝑛 2 𝛽−(𝛾−1))

= Equation 6
𝑇1 (𝛾+1)2𝑀12 𝑠𝑖𝑛 2 𝛽

Equation 3-6 are utilized for the analytical computations and are compared with
the findings of the CFD simulation. The Mach number after the oblique shock wave
was computed using the equation 7;
2
𝑀12 𝑠𝑖𝑛 2 𝛽+
𝛾−1
𝑀22 𝑠𝑖𝑛2 (𝛽 − 𝛿) = 2𝛾𝑀2 2 Equation 7
1 𝑠𝑖𝑛 𝛽−1
𝛾−1

2.2 Numerical method

The schematic representation of high-speed train moving in a tunnel is used to set the
model for the analysis. The steps for performing the CFD of the high-speed train
moving through a tunnel using the ANSYS workbench were discussed below;

Figure 1 CFD model

4
Geometry Creation: Create a geometric model of the high-speed tarin moving
through the tunnel using software such as ANSYS Design Modeler. Create the train,
tunnel and space between them. Define the fixed levitating height of the train while
varying the gap (h) between the train roof and tunnel for various conditions (0.05 < h*
< 0.50, where h* = h/C).

Mesh generation: Once the model is positioned, generate a high-quality mesh around
the entire geometry. Pay attention to the gap region as this is where flow variations
are expected because of changing gap size. A high-quality 2D mesh is created using
ANSYS Fluent Meshing, ensuring accuracy and reliability in the simulation. Use fine
mesh near the walls and around the carriage to capture boundary layer effects
accurately. Specifically, this study focuses on the fluid properties around the train, the
mesh element size around this area is set finer at 0.5mm. This refined resolution
guarantees enhanced accuracy in capturing the complex flow characteristics around
the train as depicted in Figure 3.

Figure 2 Meshing

Physical setup: The analysis assumes an ideal gas, considering density changes in
compressible fluids. The fluid is compressible and inviscid. Using an aluminium train
as a solid object, a no-slip boundary was applied. Energy equations are employed with
inlet conditions set at a residual of 0.004. An iteration of 5000 was run to get the
simulation output.

5
3. RESULTS

Table 1 displays the velocities of the train under compressible and incompressible flow
conditions along with the error percentage.

Table 1 Speed of train using compressible and incompressible flow conditions

Mach number Velocity Velocity Error (%)

(incompressible) (compressible)
(m/s) (m/s)
0.3 93.41 89.17 4.55
0.5 121.65 108.73 10.62
1 197.73 176.53 10.73
1.7 278.52 256.81 7.77
2.5 300.16 297.13 1.01

Table 1 shows the speed of a train at different Mach numbers, calculated using both
incompressible and compressible flow conditions. As the Mach number increases, the
error between the incompressible and compressible velocities also increases. This is
because the incompressible flow assumption becomes less accurate at higher speeds.
At lower Mach numbers (0.3 and 0.5), the velocities in the compressible flow condition
are consistently lower than in the incompressible flow condition, with percentage errors
of 4.55% and 10.62% respectively. The results suggest that at higher Mach numbers,
the difference in velocities between incompressible and compressible flow conditions
diminishes, with the velocities converging as the Mach number increases. Therefore,
it is important to use compressible flow calculations for high-speed train applications.

The fluid flow is analysed using CFD at a Mach number of 1.7 with a static temperature
of 290 K and a static pressure of 80000 Pa. Two places were chosen from the obtained
analytical results after and before the oblique shock wave. The ratio of pressure,
temperature and velocity of these two points were compared with the obtained
analytical result [LO3]. The pressure and velocity contour are presented in Figures 3
and 4.

6
 Figure 3 Pressure contours

Figure 4 Velocity contours

The analytical results obtained from equation 3-8 are compared with the CFD
results. It was observed that the CFD model accurately predicts the analytical results
of the train moving through the tunnel with a maximum variation of only 11.2%. The
CFD results were found to be well agreement with analytical results [LO4].

Table 2 Comparison of analytical results and CFD result

Variables Analytical CFD result Percentage

result difference (%)

7
 Wave length (β) 15.16 14 7.65
𝑀2 1.12 1.12 0
𝑃2
⁄𝑃 1.62 1.31 19.14
1

𝑇2
⁄𝑇 1.17 1.11 5.13
1
𝜌2
⁄𝜌1 1.42 1.22 14.08

Detecting the existence of oblique shock waves is essential for measuring the
fluid flow velocity in train moving through the tunnel. Utilizing oblique shock relations
aids in computing essential parameters such as shock angle, deflection angle,
downstream Mach number and pressure ratio across the shock, among other flow
characteristics. Upon comparison with CFD results, there is a generally lower
percentage discrepancy observed across all properties analysed. However, the
pressure ratio and density ratio exhibit notably higher error percentages.

4. CONCLUSION

This study investigated the fluid flow of the train moving through the tunnel under
incompressible and compressible flow assumptions using numerical and analytical
methods [LO1]. Also, developed a mathematical model of the oblique shock wave to
determine the performance parameters and create a solution for the CFD model [LO2].
The CFD simulations at Mach 1.7 align closely (within an 11.2% variation) with
analytical predictions, affirming their accuracy and validating the need for
compressible flow analysis in high-speed train aerodynamics.

5. REFERENCES

[1] Niu, J., Sui, Y., Yu, Q., Cao, X. and Yuan, Y, “Aerodynamics of railway train/tunnel
system: A review of recent research”, Energy and Built Environment, 1(4), 2020,
pp.351-375.

8
[2] Liu, S., Zhang, Y. and Cao, J, “Can High-Speed Rail Help Achieve Urban Green
Development? From a Spatial Perspective”, Frontiers in Environmental Science, 10,
2022, p.894697.

[3] Ding, F., Kareem, A. and Wan, J, “Aerodynamic tailoring of structures using
computational fluid dynamics”, Structural Engineering International, 29(1), 2019,
pp.26-39.

[4] Ferziger, J.H., Perić, M. and Street, R.L, “Computational methods for fluid
dynamics”, springer, 2019.

[5] Doyle, J.F., “Static and dynamic analysis of structures: with an emphasis on
mechanics and computer matrix methods” (Vol. 6). Springer Science & Business
Media, 1991.

[6] Bosso, N., Magelli, M., Rossi Bartoli, L. and Zampieri, N, “The influence of resistant
force equations and coupling system on long train dynamics simulations”, Proceedings
of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid
Transit, 236(1), 2022, pp.35-47.

[7] Roul, R. and Kumar, A, “Fluid-structure interaction of wind turbine blade using four
different materials: numerical investigation”, Symmetry, 12(9), 2020, p.1467.

[8] Chew, J.W. and Hills, N.J, “Computational fluid dynamics and virtual aeroengine
modelling. Proceedings of the Institution of Mechanical Engineers, Part C: Journal of
Mechanical Engineering Science”, 223(12), 2009, pp.2821-2834.

[9] Y. A. Çengel and J. M. Cimbala, Fluid Mechanics: Fundamentals and Applications,

3rd ed., McGraw-Hill Education, 2014.