Effect of Unusual Terrain on Local Wind Characteristics

Abstract

In the lowest layer of the atmosphere, the wind is slowed down by the drag effect of the numerous features on the earth’s surface such as vegetation, ground roughness, and human construction. Within this layer, wind speed generally increases with height until the top of the layer is reached, where surface drag is no longer a factor. However, mountains and valleys in the vicinity of a study site can have a significant impact on wind patterns, deviating from the typical wind characteristics seen in flat, open, suburban, or urban areas. This paper details the methodology of accounting for the wind characteristics at a site surrounded by unusual terrain in the wind tunnel.

1 Introduction

In wind tunnel tests, to measure wind-induced pressures and responses on a building model, it is important to replicate the full-scale equivalent mean wind speed profile, turbulence intensity profile, and the length scale of turbulence within the simulated atmospheric boundary layer inside the wind tunnel. A scaled test model is mounted near the center of the turntable in the test section. All major features and structures within a minimum full-scale radius of about 500 m covered by the turntable are also modeled in detail; this is to assure that their effect on the wind flows around the study site is captured accurately. Upstream of the turntable, the terrain roughness is modeled by generalized roughness distributed over the floor of the wind tunnel’s working section. In addition, specially designed turbulence generators (spires) are installed at the start of the working section of the wind tunnel. By varying the floor roughness and turbulence generators, the effects of various types of typical terrain roughness upwind of the turntable can be reproduced at the model scale. A detailed explanation of these techniques is provided by Irwin [7]. The ESDU methodology [2, 3] of ascertaining profiles for each direction is used for characterizing the oncoming turbulence and speed for each angle. However, this approach does not apply to buildings or structures that are built amid a wide variety of topographic features such as mountains, rivers, ridges, escarpments, etc. These topographic features are capable of modifying the wind profile, and this needs to be accounted for in the wind tunnel test. These local topographical features can accelerate or decelerate wind speed and turbulence depending on the situation. In reality, topographical features are complex and consist of various features similar to a rolling hill scenario. These are important for structures like bridges, tall buildings, long-span roofs, and other wind-sensitive structures where aerodynamic stability and wind-induced responses are dependent on the turbulence characteristics and wind speed at the site.

Many studies have been carried out to determine topographic multipliers for use in such scenarios. Studies on the same had started way back in the 1980s [8]. Glanville and Kwok [5] studied the measurement of topographic multipliers and flow separation in model scale and Holmes et al. [6] studied the measurement of topographic multipliers and flow separation in full scale for steep escarpments. Suresh Kumar et al. [10] studied the wind characteristics at a rough mountainous terrain to assist a bridge design. Burlando et al. [1] used numerical simulations to arrive at a parameter called altitude coefficient that provides the ratio between the mean wind velocity in complex topography and the corresponding value at sea level. Tha and Khin [11] performed numerical simulations for generating wind profiles for complex terrains. Tse et al. [12] describe how the topography-induced pitch and yaw angles affect the wind loading on structures using a series of wind tunnel tests and CFD simulations. Studies have also demonstrated the use of neural networks for the development of a global BLWT data atlas to inform the development of methods to predict topographic wind speedup for a diverse range of topography and surface roughness conditions [4, 9].

To determine the site-specific wind characteristics, especially in the absence of full-scale measurements, the effects of such topography can only be effectively determined through topographical model studies or numerical simulations in the case of complex topography.

The purpose of this paper is to examine the usefulness of wind tunnel studies in determining site-specific wind conditions for a project. This paper presents the methodology of a topographical study and provides case studies conducted at RWDI. The main goals of the topographical study are to gather data on changes in average wind speed, flow direction, and turbulence intensity at the project location and apply the same during the testing of the structure.

2 Topographical Modeling

Typical wind tunnel studies of structures are carried out at 1:200–1:500 scale models. This would hardly cover less than a 0.5 km radius of the topography around the site in the traditional sizes of the boundary layer wind tunnels, and this will not capture the effects of rolling hills/mountains for many kilometers. To account for this, a small-scale model of 1:2500–1:5000 of the project site as well as all major topographical features within a radius of 3.6–10 km radius shall be selected and tested in a boundary layer wind tunnel. The radius chosen for the topography disk is based on the size of the wind tunnel turntable as well as the extent of the topography required to be modeled in the turntable. For a 1.2 m radius turntable, the extent of topography that can be modeled in a 1:3000 scale can be 3.6 km. However, when the radius and scale increases, larger extents can be covered on the disk.

The contour model of the study site is constructed using a Computer Numerical Control (CNC) machine based on GIS information. Special care is taken to elevate and contour the approach to maintain a smooth transition to the model. The objective of the topographical study and subsequent wind simulation for the actual wind tunnel test for a structure on a larger scale of 1:200–1:500 is to ensure that the effects of the far-field topography, which could not be physically modeled in the larger-scale wind tunnel test, are captured. Since the buildings within a 0.5 km radius are present for the subsequent 1:200–1:500 scale wind tunnel tests, their impact on the local flows would be accounted for directly. Wind direction is defined as the direction from which the wind blows, measured clockwise from true north.

Initially, a topographic study at a scale ranging between 2500 and 5000 is carried out to determine the mean wind and turbulence profiles along the height at the study site. For this wind tunnel test, beyond the modeled disk, the approaching profile is determined using the ESDU method for every 20° angle of attack. This profile is used for measuring the turbulence intensity and mean wind velocity including the mountains and other topographic features in the disk. Thereafter, equivalent profiles are generated at a larger scale for the actual testing of the building, to achieve similar turbulence intensity and mean velocity as if the wind was passing through those mountains and other features.

In the topography test, measurements of the wind speed and turbulence intensity are taken at several heights ranging from 50 to 2500 m above grade. The height of interest also includes the roof height of the model, the deck height of the bridge, the top of the pylon, etc. Further, in the case of bridges, measurements are also carried out at mid-span and quarter spans of the deck as well as below the deck. The measurements are made for the full compass range nominally at 20° increments. Additional measurements are conducted at 10° increments where significant upwind topography changes occur, to capture these effects in detail. Measurements of the three components of mean wind speed and turbulence intensity are obtained using a cobra probe. This device is specifically designed for measuring wind speed and turbulence intensity.

The mean wind speeds at height $z$ (${U}_{z}$) are normalized by the free stream velocity (${U}_{g}$), which is measured at a nominal height beyond 1000 m above grade where the effect of topography ceases to influence. The ratio of the mean speed at height $z$ (${U}_{z}$) to the mean speed at 600 m over open terrain could then be calculated using Eq. (1).

$$\frac{{U}_{z}}{{U}_{600{m}_{\text{open}}}}=\frac{{U}_{z}}{{U}_{g}}\cdot \frac{{U}_{g}}{{U}_{600{m}_{\text{open}}}}$$

(1)

${U}_{g}/{U}_{600m\_\text{open}}$ can be obtained from the wind tunnel test as well if there is any flat plateau in the disk itself, otherwise, standards shall be used. Wind speed and turbulence intensities are checked at heights of interest above grade (in general—the roof height of the study building) to ensure that the wind characteristics in the large-scale wind tunnel test would be representative of the wind characteristics over the site as recorded on the topographical model. As discussed earlier, the immediate modeled surroundings as a part of large-scale test are not included in the topographical wind profile measurements. Surroundings beyond the limits of the model are simulated by spires and generalized roughness elements on the wind tunnel floor. The application of the velocity scaling factors described below accounts for, on a direction-by-direction basis, any remaining mismatch between the “target” flow parameters obtained from the topography study and the modeled flow properties at the large scale required for the wind tunnel test of the structure.

During the large-scale wind tunnel tests (for example 1:300), the reference wind speed is measured at an equivalent height which is the practical maximum height for the wind tunnel. The results of the tests are required, however, to be in terms of the 600 m above sea level velocity over open terrain, ${U}_{600m\_\text{open}}$. Therefore, a velocity scaling factor, $F$, is needed to convert the reference velocity, ${U}_{\text{ref}}$, measured over the proposed development to the appropriate value over open terrain. This factor, $F$, is based on the concept of matching peak gust velocities between the two scale models at the matching heights, which leads to the following Eq. (2), where the small-scale test is assumed to be 1:3000 scale model:

$${U}_{\text{ref}}=F\cdot {U}_{600m\_\text{open}}$$

(2)

$$F={\left\{\frac{{U}_{z}\left[1+g\cdot {I}_{z}(z)\right]}{{U}_{600m\_\text{open}}}\right\}}_{1:3000}\cdot {\left\{\frac{{U}_{\text{ref}}}{{U}_{z}\left[1+g\cdot {I}_{z}(z)\right]}\right\}}_{1:300}$$

where

$z$ :

the height of interest above the grade.

${U}_{z}$ :

the mean wind speed measured at the height of interest.

${I}_{z}(z)$ :

the turbulence intensity at the height of interest.

${U}_{600m\_\text{open}}$ :

the mean wind speed specified at 600 m above sea level over open terrain.

${U}_{ref}$ :

the mean wind speed measured at the reference height in the wind tunnel.

$g$ :

the peak factor, defined as the ratio of the peak velocity fluctuation to the r.m.s. velocity.

Velocity scaling factors ($F$) are then calculated for all wind directions for the matching heights. In addition, the final velocity scaling factors are adjusted to smooth the variations in wind direction to establish more realistic values. These values are evaluated from the mean velocity and turbulence measurements on the large-scale wind tunnel test and small-scale topography test. The actual wind tunnel test is then carried out using these factors and modified profiles and results are obtained for the study model.

3 Case Study 1

The location of the project is depicted in Fig. 1. As can be seen from the figure, the study site is surrounded by a complex topography including mountains, valleys, water bodies, and tall buildings.

Fig. 1

Location of the study building

A 1:3000 scale model including all major topographical features within a radius of at least 3.6 km from the site was tested in RWDI's 2.4 m × 1.8 m boundary layer wind tunnel. The topographical model represented in detail the influence of the major topographical features within 3.6 km of the project location. For the wind tunnel test, the influence of the terrain beyond the detailed model was represented using roughness on the wind tunnel floor and spires. Figure 2 shows the photograph of the 1:3000 scale topography model in the boundary layer wind tunnel covering 3.6 km. The 1:300 scale test with the study building (in red) and immediate surrounds for a radius of 365 m is depicted in Fig. 3. Figure 3 also shows a series of tall pink color cubes at the entrance of the tunnel mouth, which have been used to generate the wind flow characteristics required at the 1:300 scale equivalent to the measurements made at 1:3000.

Fig. 2

Topographic model at 1:3000 scale

Fig. 3

Wind tunnel test at 1:300 scale

For winds from the southwest and west as shown in Fig. 2, the model was relatively flat, making the wind tunnel measurements for those directions more affected by the upwind profile than measurements taken for other directions where the local topography would have had a greater impact.

The wind speed and turbulence intensities are measured and checked to ensure that the wind characteristics at the 1:300 scale model would be representative of the wind characteristics over the site as recorded on the 1:3000 scale topographical model. Surroundings beyond the limits of the model are simulated by non-traditional combination of spires and generalized roughness elements on the wind tunnel floor to achieve the desired wind characteristics seen in the topographical study. During the 1:300 scale studies, the reference wind speed is measured at an equivalent height of 366 m above grade in site terrain. Therefore, a velocity scaling factor, $F$, was needed to convert the 366 m reference velocity, ${U}_{\text{ref}}$, measured over the study site to the appropriate value over open terrain at 600 m, using Eq. (2).

Figures 4 and 5 show a reasonably good comparison of the normalized mean wind speed and turbulence intensities respectively at 74 m height between the 1:3000 and 1:300 scale wind tunnel test. The mean wind speed is normalized by reference height wind speeds. Remaining mismatch between the two tests has been accounted using factors during the analysis. As can be clearly seen from Figs. 4 and 5 the mean wind speed is reduced when wind is coming over the mountains (clockwise from west through east), but induced turbulence is quite high. In open directions (clockwise from east through west), mean speed is high, and turbulence is low.

Fig. 4

Normalized mean wind speed

Fig. 5

Turbulence intensity

During the analysis stage of the wind tunnel results, in the mathematical model of the wind climate, the wind speeds recorded at the nearby meteorological station are scaled up to those at the gradient height according to the local topography in the area. These wind speeds at the gradient height are not expected to change from location to location and are considered applicable to the current study. The mathematical model is then combined with the wind tunnel data to predict the frequency of occurrence of full-scale wind speeds.

4 Case Study 2

The location of the project is depicted in Fig. 6. It stands on a valley beneath a mountain and is surrounded by tall buildings.

Fig. 6

Site location of the study building

A 1:3000 scale model including all major topographical features within a radius of at least 7.3 km from the site was tested in RWDI's 5.0 m × 3.0 m boundary layer wind tunnel. The mountains and valleys, particularly to the south and southwest, can produce various effects on the wind velocity and turbulence intensity at the site. Figure 7 shows the photograph of the topography model in the boundary layer wind tunnel. The 1:300 scale test with the study building (in red) and immediate surroundings for a radius of 365 m are depicted in Fig. 8. The figure also shows a few horizontal wooden plates that are used to modify the wind profile to match the characteristics obtained from the topography test.

Fig. 7

1:3000 scale topography model

Fig. 8

Wind tunnel test at a large scale of 1:300

Before testing, measurements were made of the wind profile, without buildings and large-scale topography upwind, to confirm that the approach open terrain wind profile would be appropriate. This profile was then compared to the approach profile specified in the Code of Practice.

Measurements of the wind speed and turbulence intensity were taken at several heights ranging from 50 to 1500 m above grade. For the development and scaling of the 1:300 flow profiles measurements, nominal heights of 122 m and 170 m above site grade were of the most interest as the “matching heights” based on the roof height of the study building.

During the 1:300 scale studies, the reference wind speed was measured at an equivalent height of 457 m above site grade in site terrain. This height was the practical maximum for the wind tunnel. The results of the tests were required, however, to be in terms of the 600 m above sea level velocity over open terrain, ${U}_{600m\_\text{open}}$. Therefore, a velocity scaling factor, F, was needed to convert the 457 m reference velocity, ${U}_{\text{ref}}$, measured over the proposed development to the appropriate value over open terrain. Figure 9 shows a comparison of the peak gust speeds at both matching heights, based on the scaling factors.

Fig. 9

Comparison of simulated wind speed at 1:3000 and 1:300 scale models

Figure 9 shows that mean and peak gust measured at 1:300 and 1:3000 is in good agreement with each other. From Fig. 9 (clockwise from southeast to southwest), it is seen there is a significant reduction in wind speed due to the sheltering from the immediate mountainous terrain as is expected.

5 Conclusions

Codes recommend topographic factors which are evaluated under idealized terrains and fail to accurately predict the impact of complex topographic features such as a combination of hills, valleys, cliffs, escarpments, or ridges which can significantly affect wind flow characteristics. Numerical simulations or wind tunnel tests are quintessential in calculating the actual aerodynamics at the location. The paper reiterates the importance of conducting a topography study to accurately capture the impact of local terrain on the wind flows at the site. This study is suggested for structures in unusual terrain conditions to ensure that the site-specific wind characteristics are aptly captured in the wind tunnel test. The methodology of conducting a topography study in the wind tunnel and its application in the wind tunnel test is presented through two case studies. These case studies have also demonstrated that the results of a small-scale topography study can be translated to a large-scale wind tunnel study in good agreement.