Case 5 Help: Flared Entrance to Reduce Pressure Gradients

Case 5 simulates pressures and velocities when a train enters a tunnel with a flared entrance region. In practical applications, one purpose of such entrances is to reduce potential discomfort for passengers from pressure waves generated during train-entry. Another purpose is to reduce the amplitudes of sonic booms emitted from train exit portals because of the steepness of pressure waves generated during train-entry.

Case 5 of ThermoTun-Online can be used for both purposes, but note that three-dimensional effects invalidate some details of the predicted pressure changes caused when the train nose and tail cross the flare portal and the end of the flare. Also note that well designed entrance regions for the alleviation of sonic booms have windows as well as flares. The full version of ThermoTun enables much more complex entrance regions to be simulated.

Flared entrance regions in ThermoTun are intended to be quite long, typically between 50 m and 200 m. They should not be confused with simple rounding of the corners of the entrance portal such as might be used to improve local streamlining.


red dot Location of pressure and/or velocity output

Output Locations

One use of Case-5 simulations is expected to be the prediction of pressures experienced by passengers on trains. Another use might be to assess the potential of a flared entrance region to reduce rates of change of pressure associated with waves propagating along a tunnel towards its exit portal. In some tunnels, sonic booms are emitted into the external environment unless the train-entry event is made gradual.

Graphical output is provided as follows:

Box-1: Pressure histories in the tunnel
(1) at x = LTUN / 2
(2) at x = LFLARE / 2

Box-2: Pressure histories inside the leading and trailing coaches
(1) outside the train
(2) inside the train
NB: For unsealed trains, there is little difference between pressures inside and outside coaches. For well-sealed trains, sufficiently rapid external pressure changes are strongly damped. Large values of the sealing time constant correspond to tight seals and small values correspond to poor seals.

Box-3: Mean air velocity histories in the tunnel
(1) at x = LTUN / 2
(2) at x = LFLARE / 2

Box-4: Train speed

Text output is provided in the V01.txt file as follows:

LTUN, m: Length of tunnel (excluding the flare)

For most purposes, the meaning of the length of a tunnel is self-evident. If you want to be more precise, give the distance between the centroids of the portal cross-sections excluding the flared entrance region.

In ThermoTun-Online, the tunnel length must be at least 300m. The full version of ThermoTun allows experienced users to choose any length greater than zero. ThermoTun-Online is intended only for applications in which one-dimensional analyses can be used with reasonable confidence.

ATUN, m2: Cross-sectional area of tunnel

For most purposes, the meaning of the cross-sectional area of the tunnel is self-evident. If you want to be more precise, calculate the total volume of air in the tunnel and divide this by the total tunnel length. This is not a formal definition, but it is probably good enough.

In ThermoTun-Online, the area of the tunnel must be at least 10m2. The full version of ThermoTun allows experienced users to choose any area greater than zero. ThermoTun-Online is intended only for conventional train:tunnel applications.

With large cross-sectional areas, the assumptions of one-dimensional flow can become suspect. It is recommended that the effective diameter of the tunnel should not exceed about 5% of the length of the tunnel. (NB: The "effective diameter" is the diameter of a circle with the same cross-sectional area as the tunnel).

FTUN (f): Tunnel friction coefficient (Fanning, not Darcy-Weisbach)

The tunnel skin friction coefficient determines the resistance to airflows along the tunnel.  There are two definitions in common use:

Fanning (f ): Wall shear stress = f x ρU 2/2

Darcy-Weisbach (λ ): Wall shear stress = λ x ρU 2/8

where ρ is the density of the air and U is the mean velocity along the tunnel.

It follows from the definitions that λ=4f  so it is important to know which one is used in any particular software.  ThermoTun uses f, not λ.

The Fanning friction coefficient for real tunnels will exceed 0.002. It will usually be within the range 0.005 < f < 0.010.

DXW0, m: Numerical grid length

Values of pressure and velocity are calculated at closely spaced points throughout the tunnel system. The distance between adjacent calculation points (i.e. the numerical grid length) is either 0.25% of the total length of the tunnel or 20% of the train length (whichever is smaller). The full version of ThermoTun allows experienced users to choose any grid length greater than zero,

LFLARE, m: Length of flared entrance region

For most purposes, the meaning of the length of the flared entrance region is self-evident. If you want to be more precise, give the distance between the centroid of the portal cross-section and the cross-section of the end of the tunnel where it meets the flare.

For numerical accuracy, the length of the flared entrance region should be at least ten grid lengths and it should also be at least five tunnel diameters. The full version of ThermoTun allows experienced users to choose any length greater than zero. ThermoTun-Online is intended only for applications in which one-dimensional analyses can be used with reasonable confidence.

APORTAL, m2: Cross-sectional area of flare portal

For most purposes, the meaning of the cross-sectional area of the flare portal is self-evident. To be more precise, calculate the area of the cross-section normal to the axis of the tunnel.

The area of the flare portal must be at least as great as the area of the tunnel. The full version of ThermoTun allows experienced users to choose any area greater than zero. ThermoTun-Online is intended only for conventional train:tunnel applications.

AEND, m2: Cross-sectional area of end of flare

For most purposes, the meaning of the cross-sectional area of the end of the flare portal is self-evident. To be more precise, calculate the area of the cross-section normal to the axis of the tunnel.

The area of the end of the flare must be at least as great as the area of the tunnel and must not exceed the area of the flare portal. The full version of ThermoTun allows experienced users to choose any area greater than zero. ThermoTun-Online is intended only for conventional train:tunnel applications.

For the greatest practical benefit, the area of the end of the flare would normally be equal to the area of the tunnel. This avoids a step change in area giving rise to pressure waves when the train passes. In practice, however, construction constraints might necessitate the use of larger areas at the ends of flared regions.

The cross-sectional shapes of flared regions do not need to be the same as the cross-sectional shapes of the tunnels to which they are connected. Their aerodynamic influence is related far more strongly to their areas than to their shapes.

LZUG, m: Length of train

For most purposes, the meaning of train length is self-evident. If you want to be more precise and the train ends are strongly tapered, give the distance between the most extreme points. This is not a formal definition, but it is simple and it is probably good enough. The full version of ThermoTun allows experienced users to simulate trains comprising multiple coaches, each with any length greater than zero.

The prescribed length of the train must be at least 50m. The full version of ThermoTun allows experienced users to simulate trains comprising multiple coaches, each with any length greater than zero. ThermoTun-Online is intended only for applications in which one-dimensional analyses can be used with reasonable confidence. In reality, phenomena such as nose-entry and tail-entry to a tunnel exhibit three-dimensional behaviour that extends the period of development of pressure changes.

It is recommended that the train length should be equivalent to at least 10 tunnel diameters. This is because ThermoTun-Online is intended only for applications in which one-dimensional analyses can be used with reasonable confidence. The full version of ThermoTun allows experienced users to simulate trains comprising multiple coaches, each with any length greater than zero.

For numerical accuracy, it is preferable for the train length to be at least 1% of the total tunnel length. This is because the numerical grid length used in ThermoTun-online is approximately ½% of the tunnel length. The full version of ThermoTun allows experienced users to vary the numerical grid size within wide limits.

AZUG, m2: Cross-sectional area of train

For many purposes, the meaning of the cross-sectional area of a train is self-evident. For the calculation of airflows in tunnels, a rather special meaning is implied (irrespective of which software you use). The key need is to simulate the blockage effect of the train. For a passenger coach, for example, the effective aerodynamic area is bigger than the cross-section of the passenger compartment, but smaller than the overall structural envelope including the underbody and bogies, etc.

The area of the train must be at least 5m2 and the blockage ratio in the tunnel must not exceed 50%. The full version of ThermoTun allows experienced users to choose any area greater than zero. ThermoTun-Online is intended only for conventional train:tunnel applications.

FZUG (f) Train friction coefficient (Fanning, not Darcy-Weisbach)

The train skin friction coefficient determines the main component of resistance to airflows in the annulus between the sides of the train and the tunnel.  There are two definitions in common use:

Fanning (f ): Wall shear stress = f x ρU 2/2

Darcy-Weisbach (λ ): Wall shear stress = λ x ρU 2/8

where ρ is the density of the air and U is the mean velocity (relative to the train in this instance).

It follows from the definitions that λ=4f  so it is important to know which one is used in any particular software.  ThermoTun uses f, not λ.

If you have no experience of estimating train friction coefficients, the following values are indicative of some typical trains (but only indicative):

•  modern streamlined passenger train:             f  ≈ 0.003

•  conventional regional passenger train:           f  ≈ 0.005

•  container-like freight train:                f  ≈ 0.007

The Fanning friction coefficient for real trains will exceed 0.002.  It will usually be within the range 0.003 < f < 0.010.

KNOSE: Loss coefficient, flow from open tunnel into annulus around train

In addition to skin friction on train and tunnel surfaces in the annulus around a train, account is usually taken of local losses at the nose and tail. These influence the amplitudes of pressure waves as well as the overall resistance to flow. When the direction of airflow relative to the train is from the open tunnel into the annulus, ThermoTun allows for a stagnation pressure loss given by KNOSE x ρU 2/2, where ρ denotes the air density and U is the air velocity relative to the train in the annulus.

The prescribed value of the nose loss coefficient must be in the range 0 to 0.5. This range is sufficient for all conventional trains. The full version of ThermoTun allows experienced users to choose any value greater than zero.

STAIL: Shape factor, flow from annulus around train into open tunnel

When the direction of airflow relative to the train is from the annulus into the open tunnel, ThermoTun allows for a stagnation pressure loss given by KTAIL x ρU 2/2, where ρ denotes the air density and U is the air velocity relative to the train in the annulus.  The most suitable value of KTAIL varies strongly with the train:tunnel blockage ratio.  Instead of asking users to prescribe different values for each tunnel area, ThermoTun calculates the expected loss for a blunt-ended train and uses a shape coefficient STAIL to define the actual loss as:  STAIL x KBLUNT x ρU 2/2. 

The prescribed values of the tail shape coefficients must be in the range 0 to 1. This range is sufficient for conventional trains. The full version of ThermoTun allows experienced users to choose any value greater than zero.

TSEAL, s: Sealing time constants for leakage through coach bodies

ThermoTun calculates pressures inside trains as a consequence of leakage of air through coach bodies.

Railways commonly use a sealing time constant to characterise the leakage characteristics of coaches. Typically, a test is undertaken in which the pressure inside a coach is held at a constant value for a period and is then allowed to decay freely.

The sealing time constant is usually defined as the time required for the pressure difference to reduce to (1/e) times the original value (where e = 2.718 is the base of natural logarithms. Beware: the sealing time constant is usually determined in a static test; the corresponding value for a coach in motion can be considerably smaller. Also, it is possible for different values to exist for positive and negative pressure differences between the inside and outside of a coach.

VZUG, km/h: Speed of train

In ThermoTun-Online, the train speed is constant and is defined in km/h. The full version of ThermoTun allows accelerating, decelerating, stopping, starting and reversing. It also allows speeds to be defined in other units if required.

The prescribed value of the train speed must be in the range 50 to 400 km/h. This range is sufficient for most practical purposes.