Analysis of Natural Wind Characteristics and Review of Their Correlations with Human Thermal Sense through Actual Measurements Ki Nam Kang 1,a, Jin Yu 1,b, Doo Sam Song 2,c, Hee Jung Ham 3,d, Kook Jeong Seo 4,e, In Cheol Yun 4,f 1 Graduate School, Sungkyunkwan University, Suwon, Korea, 440-746 2 Professor, Sungkyunkwan University, Suwon, Korea, 440-746 3 Professor, Kangwon University, Hyoja-2dong, Chunchon 200-701, Korea 4 Senior engineer, Samsung electronics co., 416 Maetan-3dong, Suwon, Korea 443-742 a reomiya@korea.com, b yujin737@hamail.net, c dssong@skku.edu, d heejham@kangwon.ac.kr, e kookjeong.seo@samsung.com, f ic.yun@samsung.com ABSTRACT The purpose of this study is to propose a new cooling system that can improve the problems of conventional cooling systems such as constant and monotonous air-flow, lower temperature than the ambient conditions, and energy efficiency. To improve these drawbacks, this paper focuses on the dynamic characteristics of natural wind so these characteristics can be used as a control strategy of indoor cooling systems. In this paper, based on the field measurements, the characteristics of natural wind fluctuation are analyzed by using the Fast Fourier Transform (FFT) analysis. The strong correlations between natural wind and thermal sense of the human subject can be found by using the FFT analysis, and the relationships can be model as a slope of the low frequencies of the natural wind spectrum, β-value. It is expected that the proper selection of β-value can simulate a comfortable wind and will increase the energy efficient of cooling systems. KEYWORDS: Natural wind, 1/f Fluctuation, FFT(Fast Fourier Transform), β-value, PSD(Power Spectral Density) 1. INTORDUCTION Recently the demand for cooling is rapidly increasing in Korea due to climate change resulting from global environmental changes and the residential culture centered on apartment houses. Furthermore, there is a concern for increasing energy consumption due to the quickly rising demands for comfortableness of air conditioning with the increasing expectations on quality of life of occupants in rooms. In general, the individual needs for cooling of various consumers cannot be met due to the characteristics of conventional air-conditioning systems represented by low-temperature control and monotonous airflow. As a result, energy waste is encouraged without ensuring the thermal comfort of residents due to uneven temperature distribution of residential spaces and the severe temperature differences between indoor and outdoor. Accordingly, there are active attempts in other countries to improve thermal comfort of residents and improve energy efficiency by applying natural wind characteristics called fluctuation wind to conventional room air conditioning systems. Kuno et al. (1999) is reviewing in detail the effects of air current fluctuations on the thermal comfort of human body through experiments. Kamata et al. (1999), Zhu et al. (2006), and others are reporting on their study of the fluctuation characteristics of natural wind and mechanical wind through spectral analysis, 781
and the simulation of the fluctuation characteristics of natural wind in actual air-conditioning systems. Moreover, Shukuya et al. (1999) report on the effect of the natural wind velocity and the unique fluctuation characteristics (difference of velocity, cycles) of natural wind on the thermal comfort perception through experiments of subjects exposed to natural wind in summer. The ultimate goal of these papers is to complement the shortcomings of conventional air-conditioning systems by using the air flow fluctuation characteristics of natural wind so as to achieve comfortable and energy-saving air conditioning control. This study analyzed the natural wind in relatively comfortable mountain areas in summer through actual measurements based on the results of preceding studies. Furthermore, this study also measured the thermal sensation of subjects exposed to natural wind. The measurement data was put into a spectral analysis to investigate the fluctuation characteristics of natural wind and the correlations between the natural wind fluctuation characteristics and the thermal sensation of subjects were analyzed. Through this process, this study tried to quantitatively reveal the characteristics of more comfortable natural wind for humans. The analysis results of comfortable natural wind proposed by this study will be used to develop control logic for new air conditioning system targeted at energy saving which is the ultimate goal of this study. 2. MEASUREMENTS OF NATURAL WIND This chapter reports on the contents and results of the measurements of summer natural wind in mountain areas, and the analysis results of the correlations between natural wind and the thermal sensation of the subjects. 2.1 Contents of Measurements To analyze the characteristics of thermal comfort perception for natural wind under hot and humid environment in summer, the actual measurements were conducted in Mt. Seorak area where relatively comfortable wind can be felt in summer (July 21 to 27, 2006). The site for measurements (Figure 1) was a pension located near valley facing a mountain to the north which was an open space where there were few factors affecting wind flow. As shown in Table 1, the measurement items included weather station (temperature/humidity, wind velocity, air pressure, solar radiation, etc.) to measure the microclimate around the measurement Figure 1. Measurement scene of natural wind site, temperature/humidity, radiant temperature and the 3D wind direction/velocity to record the wind characteristics felt by subjects in detail. The measurement interval was about 30 seconds for the weather station and the temperature and humidity around subjects, and 0.1 seconds (10 Hz) for 3D wind direction/velocity due to the need for detailed analysis of the fluctuation characteristics of natural wind. In addition, to investigate the thermal sensation of subjects exposed to natural wind under the measurement environment conditions, five groups of male and female subjects participated in this experiment for one hour each from 9 am to 5 pm. The clothing ensembles of the male and female subjects were 0.7clo and 0.5clo, respectively, and their thermal sensation and comfort perception were recorded once every 30 seconds or so in sitting condition (1.0met). For the thermal sensation of subjects exposed to natural wind, the ASHRAE seven-points scale for thermal sensation (-3 to +3) was used, and for the thermal comfort perception, the ASHRAE six-points scale for thermal perception (1 to 6) was used (Table 2). As shown in Figure 1, to minimize the effect of the radiant heat on the 782
thermal sensation of subjects during the experiment, an overhang was installed so that the subjects would not be directly exposed to direct solar radiation. Table 1. Measurement items and details Measurement items Details Interval Physical elements Human subject elements Outside Temp. Radiant (globe) Temp. Humidity Wind velocity Wind direction Turbulent intensity Wind velocity/direction Direct Solar Radiation Thermal Sensation and Comfort Perception Table 2. Scales of Thermal Sensation and Comfort Perception 2.2 Actual Measurement Results for Natural Wind Data-logger+ Thermocouple SK-Sato As shown in Figure 2, the wind direction to the measurement site during the measurement period showed a nearly steady distribution, with the highest frequency of southeastern wind. Figure 3 below shows the outside thermal environment (globe temperature, outside air temperature, humidity), the wind velocity in the X-axis (east-west) direction which is the front direction from the subjects, and the comfort perception data during the experiment hours for the subjects in one day of the measurement period. The average outside temperature was 24.8 ; it was a little low at 21 in the morning and around 25 in the afternoon. The average globe temperature was 32.8. Due to intermittent rains at daybreak during the Figure 2. Distribution of wind direction measurement period, the humidity in the morning was high at about 80%, but the average humidity during the total experiment period was 54%. The average wind velocity was 0.58m/s, and the highest wind velocity was 2.95m/s (Figure 3). Most of the measured wind velocity values were 0.2m/s or 30.0s 3D Ultrasonic Anemometer 0.1s Davis Weather station 60.0s - Number of participants: Male 6 person and female 6 person in a day - 2 person (male and female) were participated for 1 hour. - Age : 22~28 - Clothing ensembles : Male : T-shirt, Panty, Short-sleeved, Beach shoes (effective clothing thermal insulation is 0.4 [clo]) Female: T-shirt, Panty, Short-sleeved, Brassiere, Beach shoes (effective clothing thermal insulation is 0.41 [clo]) - Activity level: sedentary(1.1met) - ASHRAE seven points scale for thermal sensation and ASHRAE 6 points scale for comfort perception were used. Thermal sensation scale very little little very uncomfortable comfortable uncomfortable uncomfortable comfortable comfortable 1 2 3 4 5 6 Comfort perception Cold Cool Slightly Cool Neutral Slightly Warm Warm Hot -3-2 -1 0 1 2 3 30.0s 783
lower; its frequency was about 77% of the total wind velocity measurements. Solar radiation was 1,067W/m 2 at the highest, and 338.42W/m 2 on average. Frequency(%) C S V Humidity(%) Figure 3. Measurement data for outside temperature/humidity, wind velocity, and comfort perception (July, 25) Figure 4. Frequency distribution of wind velocity Figure 5. Frequency distribution of temperature Figure 6. Correlations among air temperature, wind velocity and the comfort perception of subjects (July, 25) 784
Figure 6 shows the measurement results that represent the correlation between the natural wind velocity and air temperature changes and the comfort perception of male and female subjects between 14:00 and 15:00 on July 25 th. Even though the average outside temperature was relatively high at 26, the subjects felt comfortable when wind velocity was strong at certain moments. From this, we can see that the high momentary fluctuations of natural wind affect the comfort perception of subjects. This result is also discovered in the results of Fanger et al. (1987). Table 3 below shows the analysis results of the correlations between outside thermal environment factors (outside temperature, humidity, radiant temperature, and wind velocity) and the comfort perception of subjects. The factor with the highest correlation with the thermal comfort perception of subjects was wind velocity, followed by globe temperature, outside temperature, and humidity. Table 3. Correlations between comfort perception and thermal environment factors CSV CSV 1 Humidity Humidity -0.0407 1 Outside temperature Outside temperature 0.0531-0.7370 1 Globe temperature Wind velocity X Globe temperature -0.0835-0.7897 0.7841 1 Wind velocity X -0.0873-0.0057 0.0481 0.1364 1 The distribution of wind velocity ( X ) at the time when male and female subjects felt comfortable based on the measurement results in Figure 6 and the correlation analysis results in Table 3 is shown in Figure 7. Here, the frequency of feeling comfortable according to the momentary wind velocity was similar between male and female. The highest frequency of feeling comfortable appeared at the wind velocity around 0.4m/s. Figure 7. Relationship between comfort range and X wind velocity 3. ANALYSIS OF NATURAL WIND CHARACTERISTICS This chapter analyzes the unique fluctuation characteristics of natural wind measured in mountain areas in summer through the Power Spectral Analysis method. In particular, the review results on the effects of spectral characteristics of air current on human comfort perception is reported. 3.1 Power spectral analysis The correlations between the fluctuation characteristics of natural wind and the comfort perception of subjects can be indirectly confirmed through the records of wind velocity changes over time and comfort perception (Figure 6). However, to more accurately analyze the irregular fluctuation rhythms of natural wind, power spectral analysis through FFT (Fast Fourier Transform Analysis) is required. Existing studies (Hara et al. 1996), have reported that the fluctuation characteristics of natural wind has the 1/f characteristic, and has close relations with human comfort perception by Yamamoto (1997). Moreover, it has been reported that low frequency eddy is an important factor in human sensation by Fanger et al. (1987). Therefore, this study analyzed the natural wind measurement data with a focus on the frequency range 0.01 to 1 Hz. Accordingly, the analysis results were averaged through the power spectrum exponent (β-value) which can simply show the energy distribution of the turbulent flow of natural wind. Therefore, the energy distribution E(f) at a specific frequency range is expressed by the following equation: 785
E / β ( f ) 1 f (1) Figure 8 compares the Power Spectral Analysis results for natural wind at typical comfortable time(a) and uncomfortable time(b) based on the analysis results for the measurement data for natural wind (Figure 6) against the Power Spectral Analysis results for mechanical wind(c). Here the analysis results for mechanical wind were obtained from the data measured at 2m away from an inlet of the air flow discharged from a home air-conditioner at the height of 1.1m and the wind velocity of 2.93m/s. Comfortable natural wind(a) shows a shaper slope of 1/f from low frequency to high frequency. It has high power in the low frequency area, and the power weakens as it moves toward high frequency, showing ignorable white noises. On the other hand, for uncomfortable natural wind(b) and mechanical wind(c), the power changes slowly from low frequency to high frequency, with a small β-value. These results clearly show the fluctuation characteristics of comfortable natural wind, uncomfortable natural wind, and mechanical wind. (a) Comfortable natural wind (b) Uncomfortable natural wind (c) Mechanical wind Figure 8. The typical logarithmic power spectrum curves Figure 9 shows the correlation between β-value and average wind velocity through the natural wind data measured from this study and the frequency analysis for natural wind measurement section where male and female subject s responded relative comfortableness. The analysis results of this study showed that the comfortable area was distributed mainly at lower wind velocity, and the β-value of comfortable natural wind ranged between 1.1 and 2.0. This was almost identical to the β-value range (1.2 to 1.9) of comfortable natural wind from a similar study by Zhu et al. (2006). β Figure 9. β-value distribution of comfortable natural wind 786
4. CONCLUSIONS This study analyzed the natural wind in relatively comfortable mountain areas in summer through actual measurements in order to investigate the correlation between the fluctuation characteristics of natural wind and the human thermal comfort perception. Furthermore, this study simultaneously measured the thermal sensation of subjects exposed to natural wind. To more accurately analyze the irregular fluctuation rhythms of natural wind, a power spectral analysis through FFT (Fast Fourier Transform) analysis was conducted. These analysis results were averaged through β-value which can simply show the dynamic characteristics of natural wind. This study leads to the following conclusions: (1) Among the outside environmental factors (wind velocity, temperature, radiant temperature, humidity) that were measured, the change of wind velocity shows the highest correlation with the comfort perception of subjects. (2) The frequency of feeling comfortable was higher for both male and female subjects at a relatively low velocity of natural wind around 0.4m/s. (3) The subjects felt more comfortable at moments of high wind velocity and at points of high fluctuation even in irregular natural wind. (4) Natural wind has a sharp slope of 1/f from low frequency to high frequency. It has high power in low frequency area, and the power weakens as it moves toward high frequency, showing ignorable white noises. However, mechanical wind showed a steady power with slow slope of 1/f. This clearly shows the difference between natural wind and mechanical wind. (5) The power spectral analysis for wind velocity data measured in this study showed that the frequency characteristic value (β-value) of comfortable natural wind ranged from 1.1 to 2.0 while the uncomfortable natural wind showed a β-value lower than 1.0 which is similar to mechanical wind. ACKNOWLEDGEMENT This work was partly supported by grant 2005-0781-000 from the Samsung Electronics Co., Ltd. and grant R01-2005-000-11063-0 from the Basic Research Program of the Korea Science & Engineering Foundation. REFERENCES Byun, I. S., Seong, S. P., Shim, M. S., 1994. The Development Of Chaotic Room Air-conditioner Proceedings of the SAREK Summer Annual Conference. Hanzawa, H., Melikow, AK., Fanger, P.O., 1987. Airflow Characteristics In The Occupied Zone Of Ventilated Spaces. ASHRAE Transactions ;93(1), pp.524 539. Hara, T., Shimizu, M., Iguchi, K., and Odagiri, G., 1997. Chaotic Fluctuation In Natural Wind And Its Application To Thermal Amenity Proc. 2 nd World congress of Nonlinear Analysis, Theory, Methods & Applications, Vol. 30, No 5, pp. 2083-2813. Kitakawa, T., Nomura, T., 2003. A Wavelet-based Method To Generate Artificial Wind Fluctuation Data Journal of Wind Engineering and Industrial Aerodynamics 9,1 pp. 943-964. Kuno, S., Tanaka, M., Saito, T., 1999. A Study On Physiological And Psychological Responses In The Case Where Subjects Move To Slightly Warm Environment With Air Movement From Hot Environment J. Archit. Plann. Environ. Eng., AIJ No. 524, pp. 37-44. 787
Kuwasawa, Y., Saito, M., Kamata, M., 1999. Effects Of Fluctuating Air Movement On Thermal Comfort J. Archit. Plann. Environ. Eng., AIJ No. 526, pp. 37-42. Ouyang, Q., Dai, W., Li, H., and Zhu, Y., 2006. Study On Dynamic Characteristics Of Natural And Mechanical Wind In Built Environment Using Spectral Analysis, Building and Environment 41, pp. 418-426. Saito, M., Shukuya, M., 1999. An Analysis On The Outdoor Air Movement Providing With SUZUSISA Sensation J.Archit. Plann. Environ. Eng., AIJ No 523, pp 39-44 Shimizu, M., Hara, T., 1996. The Fluctuation Characteristics Of Natural Wind. Refrigeration ;71(821), pp.164-168 Xia, Y. Z., Niu, J. L., Zhano, R. Y., and Burnett, J., 2000. Effects Of Turbulent Air On Human Thermal Sensations In A Warm Isothermal Environment Indoor Air 2000, pp. 289-296. Yamamoto, MT., 1997. 1/f Fluctuations In Biological Systems. Proceeding of Annual International Conference of the IEEE Engineering in Medicine and Biology, Chicago, USA, pp.2692 7. 788