Pollutant transport between a street canyon and a 3D urban array as a function of wind direction and roof height non-uniformity

Description of Study Test Case

The general set-up

The following figures show schematically the general set-up of the wind tunnel experiment and the cases investigated. In general, the urban model (either with even height, marked A1, or with uneven height, marked A2) was positioned in the middle of the wind tunnel test section (Fig 2a). To simulate the oblique wind direction, the model was rotated 45 degrees in its centre (corresponding to the centre of the coordinates x,z,y). Since the length (in x-direction) of the urban model was greater than the width (y-direction) of the wind tunnel, the buildings at the upper corners of the urban model were not taken into account after the rotation, and some parts of the buildings were cut off or added according to the walls of the wind tunnel and the length of the development section.

The first reference street canyon (hereafter called as A1-R, see green rectangle in Fig 4a) was part of the urban model A1, formed by evenly spaced 8 x 4 courtyard-type buildings of constant length (L = 300 mm, i.e. 120 m at full scale) and width (W = 150 mm, i.e. 60 m at full scale) and with pitched roofs of constant height H = 62.5 mm (i.e., 25 m at full scale and corresponds to the height of the roof ridges). The height of the eaves corresponded to z/H = 0.8 (shown in sketch, Fig 3). The second (A2-R) and third (A2-L) street canyons (see green rectangles in Fig 4b) were part of the urban model A2 which had the same layout as the model A1 but had arbitrarily distributed roof heights (0.8H, H or 1.2H) along each building's wall. However, each of the non-uniform street canyons has the same mean height as that of the urban model with the constant roof height (H).

Upstream of the model, a neutrally stratified atmospheric boundary layer was simulated using roughness elements and Irwin spires in the development section of the wind tunnel. Based on the mean height of the building (H) and the free flow velocity U_ref = 6.2 ms^-1 (which was used as the reference velocity), the flow was completely independent of the Reynolds number (i.e. Re_B = HU_ref/ν = 24400, where ν is the kinematic viscosity of the air). To simulate the pollution of the street canyons, we used a 1 m long ground-level line source (red line in Fig 2b), which was positioned at the centre line of the investigated street canyons and uniformly emitted a passive gas (ethane).

Figure 2: Schematic representation of the experimental setup in the wind tunnel with reference to the wind tunnel coordinates (x,y,z). a) side view, b) top view. All dimensions are given in mm. Adapted from Nosek et al. (2023)

Figure 3: Sketch of the courtyard buldings forming the street canyon in the urban model A1.

The principal quantities of interest

All major flow (mean velocity and turbulence statistics, including momentum fluxes) and pollutant (mean and standard deviation of concentration) concentration qunatities, as well as turbulent and mean (advective) pollutant fluxes, were measured at the top (labelled T, Fig 4c) and lateral (Fig 4d) openings of the studied street canyon only for the canyons shown in green. All these qunatities are included in the data files, which are provided in the Measuring data and Results. Due to the uneven roof height, all quantities were measured at two heights in the case of the top openings. The first height was chosen at z/H = 0.6, which corresponds to the lowest street canyon wall (without taking roof pitches into account, see the side cut at A-A in Fig 4d). This height thus enclosed each street canyon of the non-uniform urban model from top. The second at z/H = 1 was chosen as the reference height for both urban models. In the case of the lateral openings, all quantitates were measured at the right (labelled R when viewed from downstream ) and left (labelled L) openings of each street canyon studied up to the height z/H = 0.6. At the top openings, the longitudinal (u) and vertical (w) velocity components were measured simultaneously, while at the lateral openings, the longitudinal (u) and lateral (v) velocity components were measured. Therefore, the vertical and lateral turbulent pollution fluxes were measured for the upper and lateral openings, respectively. The details about the experimetn and the studied cases can be found in Nosek et al. (2017)

Figure 4: Schematic representation of the studied street canyons (green rectangles) in the city models with (a) uniform (with equal height H) and (b) non-uniform height. Measurement grid for the (c) top and the (d) lateral openings for all three investigated street canyons. The red line in (a) and (b) represents the near-ground line source, and the grey contours represent the dimensionless height z/H. (d) is the vertical (xz) cut at the position y/H = -5.12. Adapted from Nosek et al. (2018)

Contributed by: Štěpán Nosek — Institute of Thermomechanics of the CAS, v. v. i.

Front Page

Introduction

Review of experimental studies

Description

Experimental Set Up

Measurement Quantities and Techniques

Data Quality and Accuracy

Measurement Data and Results