Analysis of Large Vehicle Interference on Transiting Test for Measuring Building Wind Pressure Coefficient

Abstract

As a new method for structural wind engineering research, the transiting test has good application prospects. However, in the road environment, the experimental results will inevitably be influenced by other vehicles (especially larger vehicles). In this paper, a large vehicle was set as an interference vehicle to drive in specific conditions, and the symmetrical CAARC standard model was used to investigate the large vehicle’s interference on the transiting test method for measuring building wind pressure coefficient. The results indicate that the wind pressure coefficient time history will appear as a clear up and down undulation phenomenon during the overtaking interference period instead of the single convex phenomenon. The mean wind pressure coefficient in the negative pressure area during the overtaking interference period is lower than that in the no-interference situation. The overtaking interference features are clear and easy to distinguish, and the interference duration is long (more than 6 s). In contrast, the overtaking interference when the test vehicle is overtaken is greater. For following driving, the wake of the large vehicle no longer interferes with the transiting test results after the vehicle spacing exceeds 70 m.

1. Introduction

In recent years, structural wind engineering has become a research hotspot, involving many fields. The research methods mainly include wind tunnel tests, numerical simulation, and field measurement. However, the above methods have their own advantages and disadvantages. The flow field conditions and model parameters in numerical simulations are often different from the actual situation, and tend to be idealized. The computational accuracy is usually limited by grid, incoming flow, and other conditions [1]. The aerodynamic problems cannot be completely solved by numerical simulation. Field measurement is the method that best matches the actual situation, but it is also limited by meteorology, installation conditions, measurement period, and other conditions. In addition, it is also difficult to conduct actual measurement on some large structures [2,3]. Wind tunnel testing is the most commonly used test research method, but it will be affected by the blocking effect [4], tunnel wall interference [5], and other factors. Baker and Blocken extensively discuss the use of experimental and numerical techniques in wind energy engineering applications, highlighting their major advantages and disadvantages [6,7]. Based on this, Li et al. [8,9] proposed the transiting test method, that is, using the relative wind field generated by a moving vehicle to carry out structural wind engineering research. Compared with the wind tunnel test, this method has the advantages of convenience, cost saving, and no blocking effect. At present, there are many related studies on the transiting test method. Li et al. used the transiting test method to study the aerodynamic coefficients of the CAARC high-rise building standard model [9] and the triangular prism model [8]. After comparing the results with the wind tunnel test results, it was found that the test method is feasible. Considering the interference factors, Li et al. have studied the natural wind influence [10] and the end plates influence [11]. In order to further improve the roof wind field, Li et al. [12] conducted wind profile simulation research on the roof of a test vehicle using flat plates. In addition, the transiting test method is also applied to the galloping research of iced conductors [13].

Transiting tests are carried out in the real road environment, and the stability of the flow field above the road will be disturbed by driving vehicles [14]. Overtaking behavior often occurs in the driving process, and there have been many studies on overtaking aerodynamic interference. Noger et al. [15] studied the aerodynamic interference on the overtaken vehicle when two vehicles with different sizes overtook it. Sun et al. [16] used rectangular parallelepiped models as vehicle models to analyze the changes of aerodynamic coefficients and surrounding flow fields of the two vehicles during overtaking. Kremheller et al. [17] studied the aerodynamic interference and the pressure distribution surrounding the vehicle during overtaking and passing. From the flow field diagrams surrounding the two vehicles during overtaking by Jakirlic et al. [18] and Liu et al. [19], it can be intuitively seen that the flow fields around the two vehicles will interfere with each other when overtaking occurs, and the aerodynamic forces of the two vehicles will be affected accordingly.

In addition to overtaking, following driving will also interfere with the flow field around the vehicle. Watkins et al. [20] used Ahmed vehicle model to study the aerodynamic influence of longitudinal vehicle spacing on the front and rear vehicles. From the experimental results, it can be seen that the aerodynamic force of the rear vehicle will be affected in a long vehicle spacing. Li et al. [21] studied the aerodynamic characteristics of the rear vehicle under different vehicle spacings when following driving. The results show that the aerodynamic force of the rear vehicle will be affected by the front vehicle under a long vehicle spacing. From the velocity streamline diagram, it can be found that there is mutual interference in the flow field around the two vehicles. Altinisik et al. [22] also studied the aerodynamic performance of two vehicles under different vehicle spacing. For the transiting test, the roof model is directly exposed to the road environment; overtaking behavior and following driving will inevitably affect the test results. Therefore, studying and eliminating these interferences is of great significance to improve the accuracy and reliability of the transiting test.

Combined with previous studies, the impact factors of end plate, road type, vehicle vibration, natural wind, and so on were considered. Based on previous tests, vehicle interference was studied, with vehicles divided according to the size and type of vehicles. The interference of small and medium-sized vehicles on the test road to the transit test method has been studied before. This paper focuses on the interference of large vehicles to the transit test method and uses the change of wind pressure coefficient of the CAARC standard model to reflect the interference. After qualitative and quantitative analysis of the test results, important conclusions are drawn, which improves the study of vehicle interference in transit testing. According to the setting conditions of the study on interference of small- and medium-sized vehicles to the transit test method, the investigation was carried out from the following three aspects: overtaking of test vehicles, active overtaking of test vehicles and subsequent driving. The interference was reflected by measuring the change of wind pressure coefficient of the CAARC standard building model.

2. Methodology

The CAARC standard model [23] has been widely used in wind engineering research. The model can be used not only to study wind field conditions [1] but also to examine the wind resistance of high-rise buildings [24,25]. This study adopts the CAARC standard model of 101.6 mm (L) × 152.4 mm (B) × 609.6 mm (H), and the scale ratio is 1:300. A total of 20 measuring points (5 points on each side) are arranged at a height of 2/3 H of the model, as shown in Figure 1. According to the direction of driving and incoming flow, surfaces I, II, III, and IV are the windward, inner crosswind, leeward, and outer crosswind surfaces, respectively. In addition, Nos. 3, 8, 13, and 18 measuring points are the representative measuring points on each surface.

The transiting test facility is presented in Figure 2. The test vehicle is a car, and the test model and pitot tube are installed firmly on the test platform, which is placed on the roof through the sunroof. When the test vehicle drives on a straight highway using cruise control, a relatively stable flow field can be obtained on the vehicle roof [8,9].

The wind pressure measurement system is composed of 20 high frequency dynamic pressure sensors (YMC41-D) and a multichannel USB data acquisition card (USB2805C). The high frequency pressure sensors (Figure 3) are placed in the test vehicle and connected to the roof model through flexible pipes. The wind pressure coefficient of each measuring point is expressed as follows [26,27]:

$$C_p=\frac{P_i-P_{\infty }}{0.5\rho U^2}$$

(1)

The pitot tube (as shown in Figure 2) and high frequency pressure sensors are used to measure the total pressure and static pressure of the roof flow field. Wind velocity can be calculated according to the Bernoulli equation as follows [28]:

$$\mathrm{U}=\sqrt{\frac{2\left(P_0-P_{\infty }\right)}{\rho }}$$

(2)

When Formula (2) is substituted into Equation (1), the wind pressure coefficient can be expressed as

$$C_p=\frac{P_i-P_{\infty }}{P_0-P_{\infty }}$$

(3)

The mean and fluctuating wind pressure coefficients are expressed as follows:

$$\overline{C_p}=\frac{1}{n}{{\displaystyle \int }}_{t_1}^{t_2}{C}_{p}dt$$

(4)

$$C_p^{\prime }=\sqrt{\frac{1}{n}{{\displaystyle \sum }}_{k=1}^n{\left(C_{p,k}-\overline{C_p}\right)}^2}$$

(5)

where $P_i$ is the pressure value of measuring point i, $P_0$ and $P_{\infty }$ are the total pressure and static pressure of incoming flow, respectively; $\mathrm{U}$ is the mean wind velocity of incoming flow; $\rho$ is the air density ($\rho =$ 1.225 kg/m³); $C_p$ is the wind pressure coefficient; $t_1$ and $t_2$ are the initial and end times, respectively, of the sampling duration for calculating mean wind pressure coefficients; n is the number of wind pressure data within the sampling duration; $C_{p,k}$ is the k-th wind pressure data value; and $\overline{C_p}$ and $C_{pi}^{\prime }$ are the mean wind pressure coefficient and fluctuating wind pressure coefficient.

In order to eliminate the interference of road conditions, the transiting test needs to take place on a straight and continuous road with few vehicles [8,9]. The road section from Science Avenue Station to Zhongyuan West Road Station of the Zhengzhou Bypass Expressway in China can satisfy the aforementioned conditions well (as shown in Figure 4a), so it is used as the test road. In addition, the test was conducted under natural wind conditions of less than 1.2 m/s to eliminate the natural wind effect [10,29]. The overtaking cases in this paper all occur in adjacent lanes, as shown in Figure 4b.

According to the research findings of Bhautmage and Gokhale [30], obvious differences can be found in the wakes formed by vehicles of different sizes and their influence on the flow field. With reference to the classification of vehicle size levels by Gao et al. [31] and Lichtneger and Ruck [32,33], and combining actual conditions, driving vehicles on the test road are divided into three levels: small, medium, and large (

Table 1. Classification criteria for different types of vehicles.

	Classification Criteria for Different Types of Vehicles (Unit: m)
	Small	Medium	Large
Length	1.5–5.0	5.0–6.5	$\ge$6.5
Width	1.2–2.1	2.0–3.5	$\ge$2.5
Height	1.3–2.0	2.0–3.1	$\ge$3.0
	Small	Medium	Large

). The test vehicle is a car, which is categorized as the small vehicle. A truck is selected to represent the large interference vehicle, as shown in Figure 5. The size of the entire vehicle is 6.00 m × 2.25 m × 3.05 m.

3. Results

3.1. Interference Analysis for Test Vehicle Being Overtaken

This section investigates the interference on transiting test results when the test vehicle is overtaken by the large interference vehicle. According to previous research, the lower the test vehicle velocity and the higher the relative velocity, the more obvious the overtaking interference [34]. In order to capture larger overtaking interference, the test vehicle velocity is set to a lower 60 km/h. Considering the driving safety of truck and expressway regulations, the velocity of large interference vehicle is set to 80 km/h.

Figure 5 shows the wind pressure coefficient time histories of the four representative measuring points in the large vehicle interference test; the shaded part in the figure is the overtaking interference period. Different from the test results of small and medium interference vehicles, the wind pressure coefficient time history of the large vehicle interference test is no longer a single convex phenomenon during overtaking, but an up and down phenomenon. It is found that the mean wind pressure coefficient of small and medium-sized vehicles in the overtaking interference period presents a single convex trend, while the mean wind pressure coefficient of large vehicles in the overtaking interference period presents a fluctuating trend. Apart from this, the fluctuation range of the wind pressure coefficient during the overtaking interference period is larger. Compared with No. 3 measuring point, the fluctuations of Nos, 8, 13, and 18 measuring points are more severe. In addition, it can be found that the duration of overtaking interference for large interference vehicle is very long. The mean value of wind pressure coefficient at No. 3 measuring point during overtaking interference period is slightly higher than other periods, while the mean values of wind pressure coefficient at Nos, 8, 13, and 18 measuring points during overtaking interference period are lower than other periods.

In addition, the overall distribution of the mean wind pressure coefficient curve during overtaking interference and non-overtaking interference periods was analyzed. The results are shown in Figure 6 which includes wind tunnel results [1] and simulation results of uniform flow [35]. The bulges of the mean wind pressure coefficient curves on the leeward surface (surface III) in this experiment are unclear, because the flow field on the roof of the test vehicle is relatively smooth. Hence, the overall trends of the curves are more consistent with the numerical simulation result of smooth inflow. It can be observed that the curve under overtaking interference has a large deviation from the curve under non-overtaking interference, especially in the negative pressure area. In addition, in the case of non-overtaking interference, the symmetry of the mean wind pressure coefficients on surfaces II and IV of the model is good. However, due to the interference of overtaking behavior, the mean wind pressure coefficients on surfaces II and IV do not satisfy a good symmetrical relationship. In order to quantify the degree of the above deviation, the following deviation calculation formulas are introduced [1]:

$$\mathrm{Deviation} (\%)=\frac{1}{n}{\sum }_i^n\frac{\overline{C_{pi1}}-\overline{C_{pi2}}}{\overline{C_{pi2}}}\times 100\%,$$

(6)

$$\mathrm{Absolute} \mathrm{deviation} (\%)=\frac{1}{20}{\sum }_1^{20}\left|\frac{\overline{C_{pi1}}-\overline{C_{pi2}}}{\overline{C_{pi2}}}\right|\times 100\%,$$

(7)

where n is the number of measurement points participating in the current deviation calculation (set to 5 on each surface); $\overline{C_{pi1}}$ is the mean wind pressure coefficient at measurement point i during overtaking interference period; and $\overline{C_{pi2}}$ is the mean wind pressure coefficient at measurement point i without overtaking interference.

The deviation results are shown in Figure 7, and the results of small and medium interference vehicles are used as comparison. The velocities of test and interference vehicles are 60 and 90 km/h, respectively, for the contrastive test results. When the test vehicle is overtaken by a large interfering vehicle, the deviations on surfaces II, III, and IV of the test model are all positive, which is different from the distribution law of the interference test of small and medium-sized vehicles. In addition, the deviation value on surface I is the smallest (only 3.94%) among the deviation results of the large vehicle interference tests. It shows that overtaking behavior mainly causes greater interference on surfaces II, III, IV, which is the same as the conclusion of the small and medium interference vehicle tests. Moreover, the largest deviation value is on surface II, reaching 24.44%, and the deviation values of surfaces III and IV are 12.36% and 17.64%, respectively. Therefore, compared with the distribution rules of interference tests of small- and medium-sized vehicles, the deviation values on surfaces II, III, and IV are quite different. Apart from these, since the wind pressure coefficient time history of large vehicle interference tests is the up and down undulation phenomenon instead of the single convex phenomenon during overtaking interference period (as shown in Figure 5), the total absolute deviation value is not too large, and the distribution law of deviation also changes. Although the deviation value on surface II is much larger than the other three surfaces, the total absolute deviation of all 20 measuring points for the large vehicle interference test is only 14.60%.

Figure 8 shows the fluctuating wind pressure coefficient curves with and without overtaking interference. It can be found that overtaking behavior affects the fluctuating wind pressure coefficient, and the curve has a large deviation under the overtaking interference. However, the overall trend of the two fluctuating wind pressure coefficient curves is relatively similar. The overtaking behavior of the large vehicle makes the wind pressure coefficient of each measuring point fluctuate greatly. It can be seen that the overtaking behavior of the large vehicle greatly interferes with the transiting test results.

3.2. Interference Analysis for Test Vehicle Actively Overtaking

The interference on the transiting test results when the test vehicle overtakes the large interference vehicle is investigated in this section. In order to capture larger overtaking interference, the test vehicle velocity should be set lower. Therefore, the test vehicle speed is set to 72 km/h, and the velocity of large interference vehicle is set to 60 km/h.

The wind pressure coefficient time histories of the four representative measuring points for the large interference vehicle test are shown in Figure 9, and the shaded parts are the overtaking interference period. It can be found that the fluctuation range of the wind pressure coefficient at No. 3 measuring point has increased during the overtaking interference period, and the mean value is larger than other periods. Compared with No. 3 measuring point, the wind pressure coefficient of Nos. 8, 13, and 18 measuring points fluctuate more obviously during the overtaking interference period, and the main trend has the up and down undulation phenomenon. It can be reflected that the surrounding flow field is disturbed and unstable when the test vehicle actively overtakes the large interference vehicle. However, the undulation phenomenon of wind pressure coefficient during the overtaking interference period are not as dramatic as that when the test vehicle is overtaken. In addition, the mean value of the overtaking interference period is higher than other periods, and the duration of overtaking interference is also relatively long.

The deviation result is shown in Figure 10, and the deviation result when the test vehicle being overtaken is used for comparison. The deviation value when the test vehicle is actively overtaking is positive on all four surfaces, which is the same as the rule when the test vehicle is being overtaken. The deviation values of the four surfaces when the test vehicle is being overtaken are very different, while the deviation values are relatively close when the test vehicle is actively overtaking. In contrast, the deviation values when the test vehicle is actively overtaking is smaller (except surface I). The interference on the windward side (surface I) is also large when the test vehicle actively overtakes the larger interference vehicle. The maximum deviation value (surface IV) and the minimum deviation value (surface I) are 10.22% and 5.06%, respectively. The overall absolute deviation of all measuring points is 7.77%, which is smaller than the 14.60% for the test vehicle being overtaken. Moreover, the deviation value of surface IV is the largest when the test vehicle is actively overtaking, and the deviation value of surface II is the largest when the test vehicle is being overtaken. The measuring points on the surface close to the large interference vehicle suffer the most interference.

Figure 11 shows the fluctuating wind pressure coefficients during the overtaking interference period and the non-interference period when the test vehicle is actively overtaking. The active overtaking behavior of the test vehicle caused the fluctuating wind pressure coefficients of all measuring points to be higher, indicating that the wind pressure coefficients of each measuring point had greater fluctuations at this period. The overall trends of the two curves are very similar, only in numerical differences, but the deviation degree of the two curves is much smaller than that for the test vehicle being overtaken.

3.3. Wake Interference Analysis for Following Driving

This section studies the wake interference of a large interference vehicle on the transiting test results when following driving, and aims to find a minimum vehicle spacing that is not disturbed by the wake. The driving velocities of the test vehicle and interference vehicles are uniformly set to 60 km/h. The spacing between the test vehicle and interference vehicle is measured by a laser rangefinder (TF03). The wake influence is analyzed through the variation of mean and fluctuating wind pressure coefficients.

The variation of mean wind pressure coefficient of each measuring point with the vehicle spacing for the large interference vehicle test is shown in Figure 12. In contrast, the mean wind pressure coefficients of the measuring points on surfaces II, III, and IV have a larger variation range. The mean wind pressure coefficient of each measuring point gradually tends to stabilize with the increase in vehicle spacing, and stabilizes after the vehicle spacing reaches 70 m.

Figure 13 shows the fluctuating wind pressure coefficients at various measuring points under different vehicle spacing. The fluctuating wind pressure coefficient of each measuring point is particularly large at the small vehicle spacing, especially the measuring points on surfaces II, III, and IV; the maximum value has exceeded 1.5. It can be seen that the wake of the large interference vehicle at close spacing has a huge interference on the wind field on the roof of the test vehicle. With the increase in vehicle spacing, the fluctuating wind pressure coefficient of each measuring point gradually decreases. After the vehicle spacing reaches 70 m, the curves are very close and almost coincide, and the fluctuating wind pressure coefficient reaches a steady state.

In short, the wake of the large interference vehicle will no longer interfere with the transiting test results after the vehicle spacing exceeds 70 m.

4. Discussions and Conclusions

4.1. Discussions

According to the previous analysis, whether the test vehicle is overtaken by the large interference vehicle or actively overtakes the large interference vehicle, the transiting test results will be interfered with for a long time. It is not ideal to eliminate the overtaking interference from the perspective of expanding the data calculation time. In addition, the overtaking interference caused by the large interference vehicle has characteristics that are easy to distinguish (as shown in Figure 5 and Figure 9). Therefore, for the overtaking interference caused by the large vehicle, it is recommended to record the time when the overtaking behavior occurs during the transiting test, and to ignore the wind pressure data during the overtaking interference period in combination with the data characteristics in the later data processing stage. Alternatively, in view of the low frequency of large vehicles, it is advisable to conduct the transiting test under conditions that avoid overtaking behavior with the large vehicle.

4.2. Conclusions

This study conducts an experimental investigation of the large vehicle interference that affects the transiting test method for measuring the building wind pressure coefficient, mainly from the two aspects of overtaking interference and wake interference for following driving. In the experiment, a large vehicle was set as the interference vehicle to drive under specific conditions, and the wind pressure coefficient results of the CAARC standard model were used to reflect the interference on the transiting test results. The main conclusions are as follows.

(1)

When the test vehicle is overtaken by a large vehicle, the wind pressure coefficient time histories of the measuring points on the four surfaces of the CAARC model show a clear up and down undulation phenomenon, rather than the single convex phenomenon for the small and medium interference vehicle tests. In contrast, there are larger undulation phenomenon on surfaces II, III, and IV. The overtaking interference leads to a large deviation of the mean wind pressure coefficient (especially on surface II), and the deviation values of the four surfaces are all positive. The overall deviation of mean wind pressure coefficient reaches 14.60%. The overtaking behavior also has a huge disturbance on the fluctuating wind pressure coefficient.

(2)

When the test vehicle actively overtakes a large vehicle, there is also an up and down undulation phenomenon for the wind pressure coefficient during the overtaking period, but it is not as dramatic as that when the test vehicle is overtaken. Active overtaking behavior leads to the deviation in mean wind pressure coefficient, but the deviation values of the four surfaces are not very different, and all are positive. The overall deviation of mean wind pressure coefficient is 7.77%. The overtaking behavior also interferes with the fluctuating wind pressure coefficient, but the interference degree is smaller than that when the test vehicle is overtaken.

(3)

It is recommended that the period when the overtaking behavior occurs during the transiting test be recorded, and that the wind pressure data during the overtaking interference period in the later data processing stage be ignored. Alternatively, it is advisable to conduct the transiting test without overtaking behavior of the large vehicle.

(4)

For following driving, the wake of a large vehicle greatly interferes with transiting test results. With the increase in vehicle spacing, the wake interference gradually decreases. When the vehicle spacing exceeds 70 m, the test results are no longer affected by the wake of the large vehicle. The overtaking interference features are clear, easy to distinguish, and the interference lasts for a long time.