A new atmospheric boundary layer wind tunnel simulation methodology for wind effects on large cooling towers considering wind environment variations

Abstract

The traditional atmospheric boundary layer wind tunnel model test practice employs wind fields, the flow characteristics of which are in accordance with the empirical formulae of the atmospheric turbulence presented in Codes of Practice and monographs. However, the empirical formulae presented in Codes of Practice and monographs cannot truthfully reflect the high variations of the realistic atmospheric turbulence which sometimes aggravates wind effects on structures. Based on model tests conducted in a multiple-fan actively controlled wind tunnel, it is found that most wind effects on large cooling towers change monotonically with the increase in free-stream turbulence, and the model test results are more unfavorable for a flow field of low turbulence intensity than for a flow field of high turbulence intensity with respect to the measured coherences. Thus, a new atmospheric boundary layer wind tunnel simulation methodology for wind effects on circular cylindrical structures is proposed to overcome the deficiency of the traditional atmospheric boundary layer wind tunnel model tests. The new simulation methodology includes the simulation of two realistic atmospheric boundary layer flow fields with the highest and the lowest turbulence intensities in the wind tunnel and the envelopment of model test results obtained in the two flow fields (e.g. the mean and fluctuating wind pressure distributions, the power spectral density, the coherence function, and the correlation coefficient). The superiority of the new atmospheric boundary layer wind tunnel simulation methodology over the traditional model test practice is demonstrated by comparing the model test results with the full-scale measurement data.

Keywords

cooling tower field measurement model test multiple-fan actively controlled wind tunnel turbulence intensity

Introduction

Based on wind tunnel model tests, Bearman (1968), Surry (1972), Kiya et al. (1982), and Cheung and Melbourne (1983) quantified the significant effects of free-stream turbulence on the flow around a circular cylinder at low Reynolds number (Re). Using full-scale measurements and equivalent passive wind tunnel model tests, our previous research by Cheng et al. (2017) suggested that free-stream turbulence might significantly influence the dynamic characteristics of wind effects on cooling towers at high Re. Although our previous research was based on reliable physical experiments, a drawback was recognized that the oncoming flow conditions utilized for our previous research could not be well controlled to facilitate rigorous parameter analyses. Specifically, the turbulence integral scale of the oncoming flow might be altered with the change in turbulence intensity at the full-scale condition or in a passive atmospheric boundary layer (ABL) wind tunnel, which might influence the flow around a circular cylinder according to Niemann and Hölscher (1990). In this regard, the discrepancies of wind effects observed between different turbulence intensity cases by Cheng et al. (2017) might not be the effects of free-stream turbulence alone. Fortunately, continuously varying air flow turbulence intensity can be produced while maintaining the other environmental parameters in a multiple-fan actively controlled wind tunnel according to Pan et al. (2011). Thus, the veracity of our previous research findings could be validated independently using the multiple-fan actively controlled wind tunnel simulation technique.

It was suggested by Cheng et al. (2017) that fluctuating wind pressure coefficients, power spectral densities of the wind pressure fluctuations, coherences, and the correlations between different wind pressure samples obtained on large cooling towers generally changed monotonically with the increase in free-stream turbulence. If this observation is true, the wind effects obtained in the uniform flow field with negligible free-stream turbulence (flow field A) and the traditional ABL turbulent flow field simulated based on empirical knowledge (flow field B) can define the variation ranges of the realistic wind effects, since the turbulence intensity of the realistic ABL wind approximately varies from zero to the empirical values presented in Codes of Practice (this point is validated by in situ measurements in Appendix 1). According to Cheng et al. (2017), wind effects were sometimes more unfavorable for a flow field of low turbulence intensity than for a flow field of high turbulence intensity, for example, the decrease of the free-stream turbulence strengthened the coherence between wind pressures at the windward region in low frequency range. To this end, a more reliable ABL wind tunnel simulation practice for wind effects on circular cylindrical structures should supplement the traditional model tests which are solely conducted in flow field B with the model tests conducted in flow field A. Conservative model test results can be obtained by enveloping multiple wind effects obtained in the two flow fields.

In view of this narrative, a 1:600 reduced-scale model test for wind effects on a cooling tower is conducted in the multiple-fan actively controlled wind tunnel of Miyazaki University to rigorously study the effects of free-stream turbulence in this article first. Second, based on the effects of free-stream turbulence observed, a new wind tunnel simulation methodology for wind effects on large cooling towers is proposed and presented using a case study. Finally, the superiority of the new simulation methodology is validated by comparing the model test results with the full-scale measurement data.

Experiments conducted in a multiple-fan actively controlled wind tunnel

Overview of experiments

The test is conducted in the multiple-fan actively controlled wind tunnel of Miyazaki University which is an open circuit wind tunnel with 99 independently programmed fans of 270 mm diameter at the front (see Figure 1). The test section is 2.6 m wide and 1.8 m high and the length can be adjusted up to 15.5 m. The 99 fans are arranged in a 9 wide by 11 high matrix, which can deliver flows with solely varied turbulence intensity using the power spectrum modification method according to Nishi et al. (1993, 1997, 1999), Nishi and Miyagi (1995), and Cao et al. (2001). Figure 2 shows the statistical characteristics of the five ABL wind fields simulated in the multiple-fan actively controlled wind tunnel which have the same mean wind velocity profile and turbulence integral scale profile, but different turbulence intensity profiles. According to Figure 2, the mean wind velocity profiles and the turbulence integral scale profiles simulated for Wind Fields 1–5 overlap the simulation targets based on Chinese Code GB 50009-2012 (2012) and Japanese Code Architectural Institute of Japan-Recommendations on Loads for Buildings (2004), respectively.

Figure 1.

Schematic of the 3D multiple-fan wind tunnel in Miyazaki University.

Figure 2.

Characteristics of wind fields simulated in the multiple-fan actively controlled wind tunnel: (a) mean wind velocity profiles, (b) turbulence intensity profiles, and (c) turbulence integral scale profiles.

The height of the prototype cooling tower employed for the experiment conducted in the multiple-fan actively controlled wind tunnel is 215 m. A 1:600 reduced-scale model is developed considering the requirement of the blockage ratio at the working section of the wind tunnel. There are 8 × 36 external pressure taps arranged along the meridian and circumferential directions, respectively. Re of full-scale cooling towers under design wind velocity approximates the order of 10⁸, while in the wind tunnel tests the order is 10⁵. The distortion of model test results due to the low Re can be compensated by modifying the surface roughness of the circular cylindrical model. The full-scale mean wind pressure distribution presented in Chinese Code DL/T 5339-2006 (2006) can be regarded as the simulation target. After trying several surface roughness cases, the actual aerodynamic characteristics of archetype cooling towers are successfully reproduced around the throat section of the reduced-scale model with the aid of sticking 36 ribs of 0.4 mm thickness and 5 mm spacing (see Figure 3). Since the multiple-fan wind tunnel tests are subjected to many uncertainties, our experimental results are compared with those of a previous model test conducted in the actively controlled wind tunnel by Zhao (2010) to check the validity of our experiment (see Figure 3). The comparison suggests that they agree well with each other on the windward side (0–80 degrees), but significant discrepancies are noted between them on the leeward side (80–180 degrees). The discrepancies are related to the Re difference between the two experiments.

Figure 3.

Simulation of Re effects on the reduced-scale cooling tower (e, a, and b are the rib thickness, the rib spacing, and the rib width, respectively).

Experimental results

Pressure measurement tests are conducted on the 1:600 reduced-scale model in the five ABL wind fields with flow characteristics described in Figure 2. By processing the data obtained on the model’s throat section at 24 cm height, wind effects for the five wind fields are, respectively, extracted and compared in this portion of the study. Since the mean wind velocities and the turbulence integral scales are close at 24 cm height in different wind fields (see data on the orange dashed lines in Figure 2(a) and (c)), but the turbulence intensities are 7.6%, 8.7%, 9.8%, 11.3%, and 13.1% for wind fields 1–5, respectively (see Figure 2(b)), the discrepancies observed between wind effects observed in different wind fields are entirely the effects of free-stream turbulence.

Mean and fluctuating wind pressure distributions obtained at different turbulence intensities are presented in Figure 4. In Figure 4(a), the five mean wind pressure distributions completely coincide. In Figure 4(b), the fluctuating wind pressure coefficients around the full half-circle monotonically increase with the increase in turbulence intensity, but patterns of the five fluctuating wind pressure distributions are all descending slopes with the sharp crests at 100°. It is noteworthy that before the separation point on the surface of the cooling tower, the influence of free-stream turbulence on the fluctuating wind pressure on the windward side is obvious. But behind the separation point, flow separation and vortex shedding will occur, and the vortex shedding may be the reason of the fluctuating wind pressure on the leeward side. According to Figure 4(b), the fluctuating wind pressure coefficients on the leeward side at different turbulence intensities are close, which indicates that the secondary turbulence caused by the vortex shedding is not closely related to the oncoming flow characteristics. Thus, it is possible that the vortex shedding phenomenon is unchanged at different free-stream turbulences.

Figure 4.

Mean and fluctuating wind pressure distributions: (a) mean wind pressure distribution and (b) fluctuating wind pressure distribution.

Figure 5 shows the power spectral densities obtained at different locations around the throat section. As can be seen, the power spectral densities basically follow the rule that they increase with the increase in turbulence intensity in the low-frequency region (0.1–1 Hz) at different positions around the half-circle. Besides, the power spectral densities for the 13.1% turbulence intensity case are much greater than those for the other cases in the high frequency range of 1–6 Hz (see Figure 5(a) to (c) and (f)), which is also observed by Cheng et al. (2017). This indicates that at a full-scale scenario or in an actively controlled wind tunnel, the increments in free-stream turbulence are probably small-scale turbulences, which can cause energy increase in the high frequency range for wind pressures on circular cylinders.

Figure 5.

Power spectral densities of wind pressure coefficient fluctuations: (a) 20°, (b) 30°, (c) 40°, (d) 50°, (e) 120°, and (f) 130°.

Figure 6 describes the coherence functions between the wind pressure fluctuation at 20° and those obtained at other locations. It can be observed from Figure 6(a) that the coherence between two samples both obtained in the windward region (0–60°) is stronger for a flow field of low turbulence intensity than for a flow field of high turbulence intensity in the low frequency range. This indicates that low turbulent flow features large-scale vortexes, which simultaneously influence wind pressures measured at different positions on the structure, while the high turbulent flow does not contain sufficient large-scale vortexes, and the coherences in the low-frequency range between samples are weak in this regard. However, this phenomenon is not observed between one sample obtained in the windward region (20°) and another obtained in the flow separation region (60–120°) (see Figure 6(b)). According to Figure 6(c), when samples are obtained at two locations far away, the coherence between them is comparatively low over the full frequency range, regardless of the oncoming flow turbulence. This suggests that the formation of wind pressure fluctuations in the windward region and that in the wake region are quite different in mechanism.

Figure 6.

Coherence functions: (a) 20°–50°, (b) 20°–90°, and (c) 20°–120°.

The correlations between the wind pressure fluctuation obtained at 140° and those obtained in other locations are shown in Figure 7. As can be seen, the increase in turbulence intensity can strengthen the correlation in most cases. This suggests that although the high turbulent flow does not contain sufficient large-scale vortexes, it features substantial small-scale vortexes in the flow field, which are well correlated.

Figure 7.

Correlation coefficients.

To quantify the changes taking place between datasets in Figures 5 and 6, parameters in empirical spectral formulae are identified by fitting the scattered data in Figures 5 and 6, and shown against the turbulence intensity in Appendix 2. The “dependent” versus “independent” variable plots in Figures 22 and 24 suggest that most parameters change linearly with the increase in turbulence intensity. In sum, most results from the experiments conducted in the multiple-fan actively controlled wind tunnel support our previous research finding based on more reliable data that wind effects obtained on large cooling towers change monotonically with the increase in free-stream turbulence in the turbulence intensity range [7.6%, 13.1%]. With regard to the fluctuating wind pressure coefficient, the power spectral density, and the correlation coefficient, the case with the highest turbulence intensity is the most unfavorable. However, with regard to the coherence function, the most adverse case is the one with the lowest turbulence intensity. Because all these loading characteristics are the important parameters utilized to calculate the structural responses to wind excitations for design and research purposes, a reasonable ABL wind tunnel model test practice should be based on the simulations of two flow fields with the highest and the lowest turbulence intensities that could possibly appear at the actual engineering site in the wind tunnel and the envelopment of the wind effects obtained in the two flow fields. These constitute a new ABL wind tunnel simulation methodology.

Multiple ABL wind tunnel model tests considering wind environment variations

Different from the conventional ABL wind tunnel model tests conducted in a single turbulent flow field with the empirical ABL wind characteristics (flow field B), a more reliable model test practice for wind effects on large cooling towers considering the high variability of the realistic ABL winds is based on simulations of both flow field A and flow field B, which is presented in this portion of the study.

Wind tunnel model tests are carried out in a TJ-3 ABL wind tunnel at Tongji University, Shanghai. It is a closed circuit rectangular cross-section wind tunnel, wherein the size of the test zone is 15 m in width, 2 m in height, and 14 m in length. The test wind speed can be continuously controlled in the range from 1 to 17.6 m/s. Without any passive devices, the uniform flow field with negligible free-stream turbulence (flow field A) is obtained. In uniform flow field, the non-uniformity of wind speed in the test zone is less than 1%, the turbulence is less than 0.5%, and the average flow deviation angle is less than 0.5°. Using a combination of triangular spires and roughness elements, the traditional ABL turbulent flow field for open terrain (flow field B) is simulated for the test. It is found that all the simulated flow field characteristics are close to the targets (Holmes, 2001; Simiu and Scanlan, 1996).

The 1:200 scaled pressure measurement model is made of synthetic glass, which ensures its strength and rigidity. Its prototype is a 177-m-high large cooling tower. In total, 12 × 36 measurement points are arranged along the meridian and circumferential directions, respectively. The wind speed at the model top height is regarded as the reference wind speed, which is measured by a system composed of a pitot tube and a micromanometer. The wind pressures on the tower model are obtained using a pressure measurement system composed of a DSM3000 electronic pressure scanning valve, a personal computer, and a self-programming signal acquisition system, the sampling frequency of which is 312.5 Hz. The data length at each pressure measurement point in each run is 6000.

To simulate the high Re effects, eight types of surface roughness (i.e. smooth tower, single-layer paper tape, two-layer paper tape, three-layer paper tape, four-layer paper tape, 1 mm × 0.5 mm thread, 1 mm × 1 mm thread, 1 mm × 2 mm thread) are set up on the model in both flow field A and flow field B. For each surface roughness condition, pressure measurement tests are conducted under four different wind speeds (6, 8, 10, and 12 m/s). Thus, the data for a total of 32 (8 × 4) cases are obtained in both flow fields. To avoid the end effects, the eighth circumferential section which is closest to the throat of the tower model is chosen as the characteristic section. By processing the data obtained on the eighth section, wind effects for the 32 simulation cases in both flow fields are obtained (see Figures 8 and 9). It can be found from Figures 10(a) and 11(a) that only five within the 32 cases can successfully re-simulate the actual static flow characteristics measured on Peng-cheng cooling tower (Cheng et al., 2017) in the reduced-scale model with lower Re in both flow fields, and they are selected as the candidate simulation cases. Then, comparing the fluctuating wind pressure distributions of the candidate simulation cases to the results measured on the Peng-cheng cooling tower (Cheng et al., 2017; see Figures 10(b) and 11(b)), it can be found that for both mean and fluctuating wind pressure distributions, the optimum high Re effect simulation cases are a model with a two-layer paper tape and 10 m/s wind speed for uniform flow field (see Figure 12) and a model with a four-layer paper tape and 6 m/s wind speed for traditional ABL turbulent flow field (see Figure 13). The data obtained for the two optimum cases are utilized to obtain the results of the new ABL wind tunnel simulation methodology.

Figure 8.

Wind effects for the 32 simulation cases obtained in uniform flow field: (a) mean wind pressure distribution and (b) fluctuating wind pressure distribution.

Figure 9.

Wind effects for the candidate simulation cases obtained in traditional ABL turbulent flow field: (a) mean wind pressure distribution and (b) fluctuating wind pressure distribution.

Figure 10.

Wind effects for the 32 simulation cases obtained in traditional ABL turbulent flow field: (a) mean wind pressure distribution and (b) fluctuating wind pressure distribution.

Figure 11.

Wind effects for the optimum simulation case obtained in uniform flow field: (a) mean wind pressure distribution and (b) fluctuating wind pressure distribution.

Figure 12.

Wind effects for the candidate simulation cases obtained in uniform flow field: (a) mean wind pressure distribution and (b) fluctuating wind pressure distribution.

Figure 13.

Wind effects for the optimum simulation case obtained in traditional ABL turbulent flow field: (a) mean wind pressure distribution and (b) fluctuating wind pressure distribution.

Setup of full-scale measurements

Full-scale wind effects are measured on a large cooling tower, which are employed to validate the veracity of the new ABL wind tunnel simulation methodology. A 167-m smooth-walled cooling tower was to be built in the Peng-cheng electric power station, Xuzhou, China, which is regarded as the principal tower for full-scale measurements. To its south, an adjacent cooling tower of the same size would be built, and there was an industrial complex to its west (see Figure 14). To its north and east, there was no large interfering building, but a few mounds.

Figure 14.

3D schematic of the Peng-cheng electric power station.

During the principal tower’s construction in 2009, 36 transducers were evenly installed around the tower’s throat section at 130 m height. Besides, another transducer was arranged inside a cabin, which can provide static reference pressure for measurements presented in this article.

The wind pressure transducers used are piezoresistive ones, the dimensions of which are 13 cm length, 5 cm width, and 3 cm depth. The transducers’ maximum measured value is ±2.5 kPa (corresponding to 63 m/s wind speed). Their maximum sampling frequency and precision are 100 Hz and 1/1000 maximum range, respectively. Before being installed on the prototype tower, the transducer was tested in a TJ-2 wind tunnel of Tongji University for its performance. It was found that when the oncoming flow speed was greater than 15 m/s, the noise-to-signal ratio for the transducer was kept below 5%. Besides, it was shown that the signal produced by the transducer agreed with those obtained using a high-precision electronic pressure scanivalve in the 0–6 Hz frequency domain. These proved that both static and dynamic performances of the transducer were good.

The full-scale measurement campaign started immediately after the cooling tower was constructed. From 2010 to 2015, intensive tests were performed two to three times every year. Each time, we predicted the occurrence of the strong wind scenario based on a local meteorological center’s weather forecast. Equipments were set up before the arrival of the strong wind, and 24-h simultaneous recordings for wind and wind-induced pressures were then conducted which usually continue for 1–2 weeks long. In the huge amount of data measured, those obtained from 28 November to 12 December 2011 were found to be the most effective.

According to the daily prevailing wind direction and the daily representative 10-min mean wind velocity obtained from 28 November to 12 December in 2011 (the mean wind velocities were obtained at 20 m height using anemometers arranged at the engineering site and converted to the corresponding values at 130 m height using the power law formula of mean wind profile), it was found that only wind speeds for 29 November and 8 December exceeded 12 m/s, which represent valid strong wind scenarios. However, the wind directions on the two days were quite different. On 29 November, the oncoming flow was from due east, but it was from due north on 8 December. Since some transducers installed on the tower’s north surface were found ineffective, complete half-circle wind pressure data could only be obtained on 29 November. Besides, the upstream terrain was smooth, and there were no obvious interference effects caused by neighboring cooling towers or buildings with respect to the specific wind direction of 29 November. As a result, the wind-induced pressures recorded on 29 November in 2011 are used, and an effective group of 10-min wind pressure time history samples obtained on that day with around 8% turbulence intensity at 130 m height are processed for further research.

Comparison of results

By processing the wind pressure samples obtained in sections “Multiple ABL wind tunnel model tests considering wind environment variations” and “Setup of full-scale measurements,” wind loading characteristics are obtained, respectively, for the model tests and the full-scale measurement, which are compared in this portion of the study.

Mean and fluctuating wind pressure distributions

The comparison of mean and fluctuating wind pressure distributions obtained by full-scale measurements and wind tunnel model tests is shown in Figures 15 and 16, respectively. In Figure 15(a), the model test result obtained in the uniform flow field almost overlaps that obtained in the turbulent flow field, suggesting that the free-stream turbulence has little effects on the static wind loads. Besides, the relative differences between the full-scale measurement result and the envelope of the two model test results are within the range [0%, 36.4%] in Figure 15(b), suggesting that the new wind tunnel model test technique can produce comparatively reliable static wind effects on structures.

Figure 15.

Mean wind pressure distributions: (a) model test results and (b) full-scale measurement results and the envelop of the model test results.

Figure 16.

Fluctuating wind pressure distributions: (a) model test results and (b) full-scale measurement result and the envelop of the model test results.

In Figure 16(a), the fluctuating wind pressure distribution for the model test conducted in the turbulent flow field is much greater than that for the model test conducted in the uniform flow field, so the result obtained in the turbulent flow field is the envelop of the two model test results. In Figure 16(b), the full-scale distribution is smaller than the synthesized result for the new wind tunnel simulation technique, but it is greater than the result obtained in uniform flow. This indicates that the conventional model test practice solely employing the traditional ABL turbulent flow field can produce conservative result for use, and the earlier model test practice employing the laminar flow before Jensen (1958) proposed that the flow field simulated in the wind tunnel should be similar to the actual ABL flow field might be unsafe.

Power spectral density

Figure 17 shows the normalized power spectral densities of the wind pressure fluctuations obtained at three locations around the throat sections of the cooling towers. In Figure 17(a), (c), and (e), one can hardly determine which of the two flow fields simulated in the wind tunnel is more unfavorable, so the practice of enveloping the model test results obtained in both the uniform flow and the turbulent flow to obtain the normalized power spectral density for use is reasonable. In Figure 17(b), (d), and (f), the envelops of the normalized power spectral densities obtained from the two model tests usually cover the corresponding full-scale measurement results in the full frequency domain, suggesting the reliability of the new ABL wind tunnel simulation methodology.

Figure 17.

Power spectral densities: (a) model test results at 20°, (b) full-scale measurement result and the envelop of model test results at 20°, (c) model test results at 40°, (d) full-scale measurement result and the envelop of model test results at 40°, (e) model test results at 120°, and (f) full-scale measurement result and the envelop of model test results at 120°.

Coherence function

The same situation holds true for coherence functions. In Figure 18(a), (c), and (e), one can hardly determine which of the two flow fields simulated in the wind tunnel is more unfavorable with regard to the coherences between the pressure fluctuations measured at 20° and those at other locations. This again suggests the necessity of using the new ABL wind tunnel simulation methodology. In Figure 18 (b), (d), and (f), envelops of the model test results are much greater than the full-scale measurement results, proving the conservativeness of the new ABL wind tunnel simulation methodology.

Figure 18.

Coherence functions: (a) model test results for the 20°–40° case, (b) full-scale measurement result and the envelop of model test results for the 20°–40° case, (c) model test results for the 20°–70° case, (d) full-scale measurement result and the envelop of model test results for the 20°–70° case, (e) model test results for the 20°–110° case, and (f) full-scale measurement result and the envelop of model test results for the 20°–110° case.

Correlation coefficient

The correlation coefficients between pressure fluctuations obtained at 80° and other samples are shown in Figure 19(a) and (b), and those between pressure fluctuations obtained at 140° and other samples are shown in Figure 19(c) and (d). In Figure 19(a) and (c), correlations for the model test conducted in turbulent flow are usually stronger than those for the model test conducted in uniform flow, which agrees with the finding observed in the multiple-fan actively controlled wind tunnel (see subsection “Experimental results”). Thus, the envelops of the two model test results are basically the curves obtained in the turbulent flow field, which cover the corresponding full-scale measurement results (see Figure 19(b) and (d)).

Figure 19.

Correlation coefficients: (a) model test results for the case of 80° and other positions, (b) full-scale measurement result and the envelop of model test results for the case of 80° and other positions, (c) model test results for the case of 140° and other positions, and (d) full-scale measurement result and the envelop of model test results for the case of 140° and other positions.

Conclusion

The main findings of this study concerning the new ABL wind tunnel simulation methodology for wind effects on large cooling towers are summarized below:

Experiments conducted in a multiple-fan actively controlled wind tunnel suggest independently with reliable data that free-stream turbulence significantly influences the dynamic characteristics of wind effects on cooling towers, and the dynamic wind effects generally change monotonically with the increase in free-stream turbulence in the turbulence intensity range [7.6%, 13.1%].

When the turbulence intensity is in the range [7.6%, 13.1%], the coherence between pressure samples measured in the windward region on the tower model is found to be more unfavorable for a flow field of low turbulence intensity than for a flow field of high turbulence intensity in the multiple-fan actively controlled wind tunnel. Thus, the traditional ABL wind tunnel model test practice of disregarding the variation of the free-stream turbulence in realistic ABL and solely simulating a high turbulence intensity case in accordance with the empirical formulae presented in Codes of Practice and monographs is problematic. To this end, a new ABL wind tunnel simulation methodology is proposed, which utilizes two flow fields simulated with the lowest and the highest turbulence intensities of the realistic ABL winds (flow fields A and B).

Using full-scale measurement data, it is revealed that wind effects obtained by enveloping model test results for flow fields A and B are conservative for use, and it might be unsafe to follow the traditional practice of solely using the results obtained in flow field B. The superiority of the new ABL wind tunnel simulation methodology over the traditional ABL wind tunnel model test practice is therefore validated.

The field measurement data presented in Figures 16 to 19 are based on a single 10-min sample. Considering their possible increments due to the effects of free-stream turbulence, other in situ samples are also processed and compared with the envelopes of multiple model test results. All comparisons suggest that the envelopes of multiple model test results are conservative for use. The discrepancies between the field measurement data and the envelopes of multiple model test results might be caused by the following: (1) the static reference pressure established for the full-scale measurement might hardly play the same role as the static pressure in the wind tunnel; (2) the full-scale velocity field lacks stationarity and homogeneity as compared with the wind tunnel situation; and (3) simulation of large-scale turbulence content in ABL in the wind tunnel is inadequate. Besides, large fluctuations are found for the full-scale measurement results presented in Figure 19, which might be related to the fact that the full-scale velocity field lacks homogeneity.

Footnotes

Appendix 1

Appendix 2 Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors gratefully acknowledge the financial support from the National Natural Science Foundation of China (Grant Nos 51178353 and 50978203), the National Key Basic Research Program of China (i.e. 973 Program; Grant No. 2013CB036300), and the China Postdoctoral Science Foundation (Grant No. 2016M601793).

References

Architectural Institute of Japan-Recommendations on Loads for Buildings (2004) Recommendations on Loads for Buildings. Tokyo, Japan: Architectural Institute of Japan (in Japanese).

Bearman

(1968) The flow around a circular cylinder in the critical Reynolds number regime. NPL Aero report no. 1257. Teddington: Aerodynamics Division, National Physical Laboratory.

Cao

Nishi

Hirano

et al . (2001) An actively controlled wind tunnel and its application to the reproduction of the atmospheric boundary layer. Boundary-layer Meteorology101: 61–76.

Cao

Nishi

Kikugawa

et al . (2002) Reproduction of wind velocity history in a multiple fan wind tunnel. Journal of Wind Engineering and Industrial Aerodynamics90: 1719–1729.

Cheng

Dong

Peng

et al . (2017) Effects of free-stream turbulence on wind loads on a full-scale large cooling tower. Advances in Structural Engineering21: 1437–1453.

Cheung

JCK

Melbourne

(1983) Turbulence effects on some aerodynamic parameters of a circular cylinder at supercritical Reynolds numbers. Journal of Wind Engineering and Industrial Aerodynamics14: 399–410.

DL/T 5339-2006 (2006) Code for hydraulic design of fossil fuel power plants (in Chinese).

GB 50009-2012 (2012) Load code for the design of building structures (in Chinese).

Holmes

(2001) Wind Loading of Structures. London: Spon Press.

10.

Jensen

(1958) The model-law for phenomena in natural wind. Ingenioren2: 121–128.

11.

Kiya

Suzuki

Arie

et al . (1982) A contribution to the free-stream turbulence effect on the flow past a circular cylinder. Journal of Fluid Mechanics115: 151–164.

12.

Niemann

Hölscher

(1990) A review of recent experiments on the flow past circular cylinders. Journal of Wind Engineering and Industrial Aerodynamics33: 197–209.

13.

Nishi

Miyagi

(1995) Computer-controlled wind tunnel for wind-engineering applications. Journal of Wind Engineering and Industrial Aerodynamics54: 493–504.

14.

Nishi

Kikugawa

Matsuda

et al . (1997) Turbulence control in multiple-fan wind tunnels. Journal of Wind Engineering and Industrial Aerodynamics67: 861–872.

15.

Nishi

Kikugawa

Matsuda

et al . (1999) Active control of turbulence for an atmospheric boundary layer model in a wind tunnel. Journal of Wind Engineering and Industrial Aerodynamics83: 409–419.

16.

Nishi

Miyagi

Higuchi

(1993) A computer-controlled wind tunnel. Journal of Wind Engineering and Industrial Aerodynamics46: 837–846.

17.

Pan

Zhao

Cao

et al . (2011) Analysis of aerodynamic load effect on thin plate section in multiple fan active control wind tunnel. Procedia Engineering14: 2481–2488.

18.

Simiu

Scanlan

(1996) Wind Effects on Structures. New York: John Wiley & Sons.

19.

Surry

(1972) Some effects of intense turbulence on the aerodynamics of a circular cylinder at subcritical Reynolds number. Journal of Fluid Mechanics52: 543–563.

20.

Zhao

(2010) Experimental Simulation and Aerodynamic Characteristics of Transient Wind on Structures. PhD Thesis, Harbin Institute of Technology, Harbin, China (in Chinese).