Wind environment around an airport terminal building

Application Challenge 4-01 © copyright ERCOFTAC 2004

Overview of Tests

Wind tunnel tests were conducted by BMT Fluid Mechanics Ltd, on behalf of WS Atkins Consultants. A subset of this data is presented in this Application Challenge, namely the runway wind environment measurements.

These include mean and RMS measurements of each of the three wind velocity components, at 4 heights and 7 locations along the runway, for 5 incident wind directions $30^{o}$ apart. Measurements were made using standard hot-wire anemometers with cross-wire probes. DOAPs for this set of experiments are mean velocity defect and turbulence intensity excess, over the runway in the wake of the terminal building.

The scale model reproduced the terminal building geometry in considerable detail, including the various major openings of the building at ground-floor level, and the major details of the internal layout at ground level. These features were included in the model for purposes of modelling the detailed ground-level wind environment within as well as around the building, though the focus in this Application Challenge are measurements taken in the wake of the building.

For pictures of the main terminal model building see Figure_tunnel_N.gif, Figure_tunnel_E.gif and Figure_tunnel_SW.gif.

Surrounding buildings in the immediate vicinity were also represented, though with a coarser level of detail. figure_c.gif and Figure_tunnel_iso_NE.gif show the model (terminal building plus salient surrounding buildings) inside the wind tunnel. figure_d.gif and figure_e.gif show four elevation drawings of the exterior of the terminal building. Building B (see figure_b.gif) was not included however, due to wind tunnel width constraints. In fact, some measurements were taken with and without building B as a sensitivity test; these indicated that the effect of building B on the wind environment around the building was negligible, within the experimental error margin.

The external geometry of the main and surrounding buildings is however well-defined and is supplied in detailed 3-D CAD files (complete_model_cad_files.zip), containing files in .IGS and .DXF format). These CAD files were created by Neil Leggatt (Atkins) based on drawings and detailed architectural CAD models of the building. The CAD models contain nearly all the detail included in the wind tunnel (with the exception of a series of low level ‘booths’ east of the main building which are omitted). For screenshots of the geometry represented in the CAD model see Figure_CAD.gif, Figure_CAD_N.gif, Figure_CAD_S.gif, and Figure_CAD_U.gif.

Figure_CAD.gif Isometric view of the main and surrounding model buildings included in the CAD model.

The area surrounding the actual terminal building consists mainly of open airfield, with some buildings and parked aircraft north of the building; the effect of the surrounding area has been represented in the tunnel with the use of ground roughness elements to achieve an equivalent ground aerodynamic roughness, $z_{0}$ , in the range of 0.03 to 0.1m (in full scale).

The dimensions of the wind tunnel were 4.8m wide, 2.4m high, with a working section 15m long. The wind tunnel tests were conducted at a scale of $\alpha =L_{m}/L_{f}=1/200$ . This is a typical scale reduction used in wind tunnel studies of atmospheric flows. However, with such a length scale reduction it is usually not possible to match the Re number in the wind tunnel to that in full scale. In this case, a wind speed ratio of $V_{m}/V_{f}=1$ was probably assumed and thus ${\mbox{Re}}_{m}/{\mbox{Re}}_{f}=200$ . Not using strict dynamic similarity is justified by the standard, well‑established argument of 'Re number independence' [Ref. 3]. According to this, geometrically similar flows are similar at all sufficiently large Re numbers, i.e. above a threshold when the flow becomes turbulent and the gross structure of the turbulence becomes similar over a very wide range of Re numbers. Sampling/averaging times for the velocity measurements were sufficiently long (of the order of a few minutes) to ensure that the ensemble average of the quantities did not change over time (there should be no effect on velocities measured and on DOAPs in particular).

NAME	GNDPs	PDPs (problem definition parameters)	MPs (measured parameters)
	Re	Wind direction	detailed data	DOAPs
EXP 1 runaway wind environment	$5x10^{6}$ (full scale) $2.5x10^{4}$ (wind tunnel)	180, 150, 210, 120, 240	$U_{x},U_{y},U_{z},\sigma _{u},\sigma _{v},\omega _{w}$	Mean velocity defect, turbulence intensity excess

Table EXP-A Summary description of all test cases

	MP1 Normalised mean wind velocities (nd)	MP2 $\sigma _{x}/\sigma _{x0},\sigma _{y}/\sigma _{y0},\sigma _{z}/\sigma _{z0}$ (nd)	Boundary data: inlet mean velocity profile
EXP 1	exp1.dat	exp1.dat	inflow a.dat

Table EXP-B Summary description of all measured parameters and available datafiles

Test Case EXP-1

Description of Experiment

Runway wind environment measurements: mean and turbulence measurements of each of the three wind components (x: along the runway, y: across the runway, and z: vertical), at 4 heights and 7 locations along the runway, for 5 incident wind directions $30^{o}$ apart (conventions defined in figure_a.gif).

The local cartesian coordinate systems used to report the results is defined as follows: z=0 is at ground level, y=0 is at the centreline of the runway, and the x position of the origin along the runway varies for each of the runs. The origin of the coordinate system is different for each wind direction run: x=0 is placed at the intersection point of straight lines drawn from the centroid of the building parallel to each of the wind directions, with the runway centreline. For each wind direction the measurement points were placed along the runway on either side of the point directly downstream of the building (i.e. the moving origin defined above). The position of the origins for each of the runs is clearly illustrated in figure_a.gif: points A, B, C, D, and E are the origins for cases q = $120^{o}$ , $150^{o}$ , $180^{o}$ , $210^{o}$ , and $240^{o}$ respectively. A coordinate triad is also shown on the plot, to define the direction of positive x which is eastwards along the runway.

Measurements were made using standard hot-wire anemometers with cross-wire probes (all three-components were measured simultaneously). Wind direction is the only problem definition parameter (PDP) for this set of experiments. Both mean and turbulence quantities are presented non‑dimensionalised with respect to values measured at the same location but in the empty tunnel (subscript o). The values measured in the empty tunnel are included in inflow_a.dat.

DOAPs for this test case are the velocity deficit, δ (in the cross-runway direction, $y$ , for velocity component V), and the turbulence intensity excess, $I$ . These are used to assess the effect of the building on aircraft landing and takeoff.

These are defined as:

$\delta ={\frac {V_{0}-V}{V_{0}}}=1-{\frac {V}{V_{0}}}$

${I}-{I_{0}}$

where $I$ is the turbulence intensity, which for the wind tunnel data may be approximated as:

${I}={\frac {\sigma _{x}+\sigma _{y}+\sigma _{z}}{V_{0}}}={\frac {\sigma _{x}}{\sigma _{x0}}}{\frac {\sigma _{x0}}{V_{0}}}+{\frac {\sigma _{y}}{\sigma _{y0}}}{\frac {\sigma _{y0}}{V_{0}}}+{\frac {\sigma _{z}}{\sigma _{z0}}}{\frac {\sigma _{z0}}{V_{0}}}$

To calculate the turbulence excess parameters from the experimental measurements, data from the empty tunnel ‘base’ case are required and these are given in inflow_a.dat.

Boundary Data

Information on the incident wind profile (mean and turbulence wind velocity variation with height) is provided in inflow_a.dat (this wind profile was measured at the building location but in the empty tunnel, in the absence of the model). The inflow boundary layer target roughness was zo=0.003m, appropriate for an open airfield consisting of grass and paved areas. This was approximated in the tunnel with a uniform distribution of roughness elements upstream of the model and a fence with an arrangement of spires at the entrance of the test section, with the remainder of the wind tunnel floor surrounding the model was covered with short‑pile carpet. The profiles achieved in the tunnel were measured and were reasonably close to those given in ESDU [Ref. 4]for an open airfield with surface roughness of zo=0.003m (see figure_g.gif).

The model building had several small openings (for the purpose of modeling the flow inside the building, and some of the flow can leak though the building. figure_f.gif is a plan view of the ground floor interior showing the main openings. Figure_f_east.gif, Figure_f_west.gif, Figure_f_south.gif give further information on the size and the locations of the openings. Nonetheless, for this test case the effect of these openings can be omitted without significantly affecting the DOAPs, since the flow characteristics over the runway should be governed by the overall size and shape of the building.

Measurement Errors

The data are expected to be accurate to at least within ±10%, based on previous experience and calibration checks for cross-wire probes.

Measured Data

inflow_a.dat : incident wind profile measured in wind tunnel, compared to an ESDU profile (ASCII file; headers: Re, wind direction; columns: z, $U_{y}(z)/U_{y}(10)$ , $\sigma _{x}/U_{y},\sigma _{y}/U_{y},\sigma _{z}/U_{y}$ , plus ESDU profile information)

exp1.dat : (ASCII file; headers: wind direction; columns: $x,z,U_{x}/U_{xo},\sigma _{x}/\sigma _{xo},U_{y}/U_{yo},\sigma _{y}/\sigma _{yo},U_{z}/U_{y0},\sigma _{z}/\sigma _{zo}$ )

References

[Ref. 3] Snyder WH (1981) Guidelines for fluid modelling of atmospheric diffusion, USEPA Report EPA-600/8-81-009. EPA Office of Air Quality Planning and Standards, Research Triangle Park, NC.

[Ref. 4] ESDU Data Item 9203 'Computer program for wind speeds and turbulence properties: flat or hilly sites in terrain with roughness changes'.

Contributors: Athena Scaperdas; Steve Gilham - Atkins

Site Design and Implementation: Atkins and UniS