- Email: [email protected]
Overall thermal sensation, acceptability and comfort
ARTICLE IN PRESS
Building and Environment 43 (2008) 44–50 www.elsevier.com/locate/buildenv
Overall thermal sensation, acceptability and comfort Yufeng Zhanga,, Rongyi Zhaob a
Department of Architecture, South China University of Technology, Wushan, Guangzhou 510640, PR China b Department of Building Science, Tsinghua University, Beijing 100084, PR China Received 6 July 2006; received in revised form 8 November 2006; accepted 27 November 2006
Abstract The relationships between overall thermal sensation, acceptability and comfort were studied experimentally under uniform and nonuniform conditions separately. Thirty subjects participated in the experiment and reported their local thermal sensation of each body part, overall thermal sensation, acceptability and comfort simultaneously. Sensation, acceptability and comfort were found to be correlated closely under uniform conditions and acceptable range ran from neutral to 1.5 (midpoint between ‘Slightly Warm’ and ‘Warm’) on thermal sensation scale and contained all comfortable and slightly uncomfortable votes on thermal comfort scale. Under non-uniform conditions overall thermal acceptability and comfort were correlated closely. However, overall thermal sensation was apart from the other two responses and non-uniformity of thermal sensation was found to be the reason for the breakage. Combining the effects of overall thermal sensation and non-uniformity of thermal sensation, a new thermal acceptability model was proposed and the model was testified to be applicable to uniform and non-uniform conditions over a wide range of whole body thermal state from neutral to warm. r 2007 Elsevier Ltd. All rights reserved. Keywords: Thermal sensation; Thermal acceptability; Thermal comfort; Non-uniform environment; Non-uniformity of thermal sensation
1. Introduction With the requirements of energy saving, more and more attention is paid recently to the study on thermally nonuniform environment. Assessment of non-uniform environment is a highlighted problem for well-designed nonuniform environment practice in buildings. For the assessment of uniform and steady thermal environment, the known indices, such as PMV, ET* and SET, which predict human thermal sensation by environmental parameters and personal informations, are widely accepted. However, there is no universal index to evaluate thermally non-uniform environment, and overall (whole body) thermal sensation [1–4], overall thermal acceptability [5–8] and overall thermal comfort [9–12] were used separately by different researchers. It would be useful to understand the relationship between overall thermal sensation, acceptability and comfort under non-uniform environment. Corresponding author. Tel.: +86 20 39851485; fax: +86 20 87110164.
E-mail address: [email protected] (Y. Zhang). 0360-1323/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.buildenv.2006.11.036
The relationship between thermal sensation and acceptability was firstly clarified by Fanger [13], who defined the dissatisfied based on the experimental results by Gagge et al. [14] as those who vote ‘Cool’ or ‘Cold’, ‘Warm’ or ‘Hot’. This definition was confirmed by Berglund [15] through comparison with the responses obtained by directly asking subjects whether they find the thermal conditions acceptable or unacceptable. As the relationship was derived and confirmed under uniform conditions, its validity under non-uniform conditions remains untested. The relationship between thermal comfort and acceptability was investigated by Berglund [15]. He compared the effect of temperatures that deviate from those of optimum comfort assessed by percent comfortable [16] with the one by thermal acceptability [17] and found that they were quite similar, which indicates that the thermal comfort votes falling in comfortable or slightly uncomfortable range were perceived by the subjects as acceptable. The comparison was conducted under uniform environment and the one for non-uniform environment remains vacant. The known relationships between thermal sensation, acceptability and comfort were derived qualitatively under
ARTICLE IN PRESS Y. Zhang, R. Zhao / Building and Environment 43 (2008) 44–50
uniform environment. The purpose of the present study was to investigate quantitatively the relationships between overall thermal sensation, acceptability and comfort under uniform and non-uniform environments and to develop a new thermal acceptability model applicable to both environments. 2. Experimental methods The experiment was carried out in the climate chamber in Tsinghua University. The chamber was used to achieve a thermally uniform environment and the ambient air temperature in the chamber was maintained with a precision of 0:2 C. Personalized ventilation system was used to produce non-uniform environment by supplying local cooling airflow to three sensitive segments of human body—face, chest and back—separately. Three room temperatures, ranging from neutral to warm, were chosen and for each room temperature, three local cooling target temperatures (target temperature means the air temperature at the center of cooling body part surface), ranging from neutral to slightly cool, were studied in the present study (Table 1). Relative humidity was kept at 40%, and air velocity was less than 0.1 m/s in the room air. Air velocity at the outlet of local cooling airflow was maintained at 1 m/s. Each test consisted of half-an-hour exposure to uniform condition and half-an-hour exposure to non-uniform condition. The ambient room temperature was maintained Table 1 Experimental conditions Room temperature (1C) Target temperature (1C)
28–32–35 22–25–28
a Hot
+2
Warm
+1
Slightly warm
+1
0
the same and local cooling airflow was supplied only when the exposure to non-uniform condition started. Thirty randomly selected Chinese students, with a normal range of age, height and weight, participated in the experiment of all conditions and the total duration for each subject was 27 h. All the subjects wore only shorts during the experiment to keep the same clothing insulation for the three cooling body segments. The sequence of presentation was balanced among the subjects. Three subjects participated in the test at one time and remained sedentary throughout each exposure. Conversation was permitted but the subjects were not allowed to exchange views concerning the thermal environment. Subjects reported their local thermal sensation of each body part, overall thermal sensation, overall thermal acceptability and overall thermal comfort simultaneously at each voting time and three times in the last 10 min for each exposure. Thermal sensations were reported on ASHRAE 7-point scale (Fig. 1a). A visual-analogue scale indicating acceptability, originally developed to evaluate indoor air quality by Gunnarsen and Fanger [18], was used in the present study (Fig. 1b). A thermal comfort scale developed by Zhang [11] was applied in the present study to force subjects to make a clear determination about whether their perceived state falls in the category of ‘Comfortable’ or ‘Uncomfortable’ (Fig. 1c). 3. Results and discussion Shapiro–Wilk’s W test was applied and the results show that human responses obtained in all conditions were normally distributed. They were therefore analyzed using repeated-measure ANOVA. Compared with independentmeasure design, repeated-measure design with balanced order of presentation is a more efficient approach, which
b
+3
Neutral
-1
Slightly cool
-2
Cool
-3
Cold
0 0
c +2
Very comfortable
+1
Comfortable
Clearly acceptable
Just acceptable
0
Just comfortable
0
Just uncomfortable
Just unacceptable -1
Thermal sensation scale
-1
45
Clearly unacceptable -2
Thermal acceptability scale
Uncomfortable
Very uncomfortable Thermal comfort scale
Fig. 1. Voting scales: (a) thermal sensation scale; (b) thermal acceptability scale and (c) thermal comfort scale.
ARTICLE IN PRESS Y. Zhang, R. Zhao / Building and Environment 43 (2008) 44–50
1.2
0.8 Overall thermal comfort
Overall thermal acceptability
46
0.4
0.0 TA = -0.41*TS + 0.58 R2 = 0.96, p < 0.001 -0.4 -0.5
0.0 0.5 1.0 1.5 Overall thermal sensation
2.0
0.6
0.0 TC = -0.79*TS + 1.01 R2 = 0.96, p < 0.001 -0.6 -0.5
0.0 0.5 1.0 1.5 Overall thermal sensation
2.0
means the same level of accuracy can be achieved by using fewer subjects. Thirty subjects in repeated-measure design with balanced order of presentation, which was used successfully in the study of environmental effects on simulated office work in a closely controlled environment [19], was applied in the present study to attain the test accuracy efficiently. It was found that human responses reached steady state within 20 min during each exposure ðp40:05Þ. If not mentioned specifically, all responses reported below are steady state responses. 3.1. Uniform environment
Overall thermal acceptability
Fig. 2. Overall thermal acceptability, TA, and overall thermal comfort, TC, as functions of overall thermal sensation, TS, under the uniform conditions.
0.8
0.4
0.0
-0.4 -0.4
TA = 0.51*TC + 0.07 R2 = 0.99, p< 0.001 0.0 0.4 0.8 Overall thermal comfort
1.2
Fig. 3. Overall thermal acceptability as a function of overall thermal comfort under the uniform conditions.
Human responses to the uniform conditions were analyzed and the results are shown in Figs. 2 and 3. Each point in the figures represents the mean vote of all subjects’ responses for each condition. It can be seen that a straight line fits the data well for each figure ðR2 40:96Þ. Under the uniform conditions, overall thermal sensation, acceptability and comfort are correlated with each other closely. Thermal sensation mean vote of 1.5 corresponds to thermal acceptability mean vote of 0 and thermal comfort mean vote of 0:2, that is to say, acceptable range runs from neutral to 1.5 on thermal sensation scale and contains all comfortable and slightly uncomfortable votes on thermal comfort scale, which confirms the relationships found by others [13,15] under uniform conditions.
similar with the one obtained under the uniform conditions. The scattered points in Fig. 4 indicate that it is difficult to define acceptable or comfortable range only by whole body thermal sensation vote as each thermal sensation vote corresponds to a wide range of thermal acceptability and comfort. For instance, thermal neutrality corresponds to the acceptability range of (0.07, 0.54) and the comfort range of ð0:02; 0:97Þ. Thus the relationship proposed by Fanger is not suitable for non-uniform environment. Except for overall thermal sensation, there should exist other important factors influencing thermal acceptability and comfort under non-uniform conditions.
3.2. Non-uniform environment
3.3. Non-uniformity of thermal sensation
Human responses to the non-uniform conditions were analyzed and the results are shown in Figs. 4 and 5. Each point represents the mean vote of all subjects’ responses for each condition. Compared with the linear relationships obtained in the uniform conditions, no linear relationships exist between overall thermal sensation and acceptability and comfort under the non-uniform conditions (see Fig. 4). Subjects feel more uncomfortable and unacceptable with the non-uniform environment than the uniform one while their overall thermal sensations maintain the same. The linear relationship between overall thermal acceptability and comfort is retained well ðR2 ¼ 0:98Þ under the nonuniform conditions (see Fig. 5) and the function is very
McNall and Biddison [20] studied thermal sensation and comfort of sedentary persons exposed to asymmetric radiant fields and found that it was ‘uneven body temperature’ which caused the thermally neutral subjects participating in the Hot Wall series to have a significantly lower probability of feeling comfortable than the subjects in the uniform conditions, where the ‘uneven body temperature’ means one side of the body feels warmer (or cooler) than the other. ‘Uneven body temperature’ was investigated by using an additional questionnaire in the present study and it was found that 97% of the subjects perceived obvious non-uniformity of thermal sensation between different body parts during the non-uniform
ARTICLE IN PRESS 0.8 Overall thermal comfort
Overall thermal acceptability
Y. Zhang, R. Zhao / Building and Environment 43 (2008) 44–50
0.4
0.0
47
1.2 0.8 0.4 0.0 -0.4
-0.4 -1
1 2 0 Overall thermal sensation Non-uniform
-1
0 1 Overall thermal sensation Uniform
2
0.8 0.4 0.0 TA = 0.52*TC + 0.06 R2 = 0.98, p < 0.001 -0.4 -0.4
0.0 0.4 0.8 Overall thermal comfort
1.2
Overall thermal acceptability
Overall thermal acceptability
Fig. 4. Overall thermal acceptability and comfort as functions of overall thermal sensation under the non-uniform conditions.
0.8 TA = -0.27*TSD + 0.57 R2 = 0.89, p < 0.001
0.6 0.4 0.2 0 -0.2 0.0
Fig. 5. Overall thermal acceptability as a function of overall thermal comfort under the non-uniform conditions.
3.4. A new thermal acceptability model Subjects evaluate thermally non-uniform environment based on their perceptions of overall thermal sensation and non-uniformity of thermal sensation. Supposing these two perceptions are independent from each other, a new
2.0
Fig. 6. Overall thermal acceptability as a function of the maximum thermal sensation difference between body parts, TSD , when in overall thermal neutrality.
1.4 Overall thermal comfort
exposures. Non-uniformity of thermal sensation may be the reason for the scattering of the points in Fig. 4. Considering the strongest feeling comes from the difference between the coolest and the warmest body parts, the maximum thermal sensation difference between body parts was chosen to represent the non-uniformity of thermal sensation. Taking the responses obtained when the subject’s whole body thermal sensation votes close to neutral (mean vote of overall thermal sensation falls in ð0:2; þ0:2Þ), the relationships between the maximum thermal sensation difference between body parts and overall thermal comfort and acceptability were analyzed and the results are shown in Figs. 6 and 7. When the nonuniformity is small, subjects perceive the non-uniform environment to be similarly comfortable and acceptable as the uniform one. With the increase of non-uniformity, overall thermal comfort and acceptability go down apparently and linearly ðR2 40:88Þ. Non-uniformity of thermal sensation well explains the breakage of the relationship between overall thermal sensation and acceptability and comfort.
0.4 0.8 1.2 1.6 The maximum thermal sensation difference between body parts
TC = -0.55*TSD + 1.01 R2 = 0.88, p< 0.001
1 0.6 0.2 -0.2 0.0
0.4 0.8 1.2 1.6 The maximum thermal sensation difference between body parts
2.0
Fig. 7. Overall thermal comfort as a function of the maximum thermal sensation difference between body parts when in overall thermal neutrality.
thermal acceptability model was proposed: TA ¼ TA1 þ TA2 ,
(1)
where TA is overall thermal acceptability, TA1 is the uniform term and TA2 is the non-uniform term. The uniform term is a function of overall thermal sensation TS, and the function was obtained by linear regression of the data obtained under the uniform
ARTICLE IN PRESS Y. Zhang, R. Zhao / Building and Environment 43 (2008) 44–50 Predicted thermal acceptability
48
way under uniform environment and non-uniform environment. The effect of non-uniformity of thermal sensation on thermal acceptability obtained under thermal neutral conditions is valid for warm conditions. The acceptable range according to the new thermal acceptability model can be expressed as
0.6 TS (0,0.5) TS (0.5,1.5) 0.3
0.0 R2 = 0.91 p < 0.001 -0.3 -0.3
0.0 0.3 Measured thermal acceptability
0:41 TS þ 0:58 0:27TSD 40.
0.6
4. Discussion
Predicted thermal acceptability
Fig. 8. Validation of the new thermal acceptability model in the nonuniform conditions that vary from neutrality.
0.6
0.3
0 R2 = 0.98 p < 0.001 -0.3 -0.3
0 0.3 Measured thermal acceptability
0.6
Fig. 9. Validation of the new thermal acceptability model in the uniform conditions.
conditions (see Fig. 2): TA1 ¼ 0:41 TS þ 0:58.
4.1. Thermal sensation and comfort The present study indicates that the mean vote obtained by thermal sensation scale is in good agreement with the one by thermal comfort scale and thus it is reasonable to use only one of the two scales or combine them together into a single scale to assess the thermally uniform environment. A good example for the latter one is the well-known Bedford scale (see Fig. 10), which confounds thermal sensation and comfort and employs the terms ‘too’ and ‘much too’ indicating that the observer is not satisfied with his thermal environment [21]. McIntyre [22] indicated that the results obtained by the Bedford scale and the ASHRAE 7-point scale may be compared directly with each other, which confirms the findings obtained in the present study. However, the close relationship between thermal sensation and comfort does not exist under the +3
Much too warm
+2
Too warm
+1
Comfortable warm
(2)
The non-uniform term is a function of the maximum thermal sensation difference between body parts TSD . The function was obtained by linear regression of the data under the non-uniform conditions while whole body thermal sensation closes to neutral (see Fig. 6): TA2 ¼ 0:27 TSD .
(4)
(3)
In order to validate the new thermal acceptability model in the non-uniform conditions that vary from neutrality and the uniform conditions, all responses obtained in the present study were collected and the predicted overall thermal acceptability calculated from the actual overall thermal sensation and non-uniformity of thermal sensation was compared with the actual overall thermal acceptability (see Figs. 8 and 9). The responses in neutral–cool range were not included in the analysis as Eq. (2) was obtained in neutral–warm conditions. It can be seen that the predicted values are highly correlated with the measured values ðR2 40:91; po0:001Þ. The new thermal acceptability model is suitable for both neutral and warm, uniform and non-uniform conditions. Subjects evaluate the acceptability of thermal environment from two aspects separately: whole body thermal sensation and non-uniformity of thermal sensation. Whole body thermal sensation affects thermal acceptability in the same
0
Comfortable
-1
Comfortable cool
-2
Too cool
-3
Much too cool
Fig. 10. Bedford scale.
ARTICLE IN PRESS non-uniform conditions as shown in the present study, which indicates that the Bedford scale is not applicable to the assessment of thermal comfort in non-uniform environment. The best way is to directly ask subjects whether or not they feel comfortable or to report their perceptions on the thermal comfort scale. Thermal neutrality for a person is defined as the condition in which the subject would prefer neither warmer nor cooler surroundings. Usually thermal neutrality is regarded the same as thermal comfort, which is confirmed by the present study (see Fig. 2) under the uniform conditions. Under the non-uniform conditions, thermal neutrality unnecessarily correspond to thermal comfort. The more the non-uniformity of thermal sensation perceived by the subjects, the more the discomfort they report even when they are in overall thermal neutrality (see Fig. 7). This is confirmed by the experimental results obtained by Olesen and Nielsen [23] and Melikov et al. [5]. They studied human responses to local cooling with air jets in warm conditions and found that the air jet velocity preferred by the subjects was not the one corresponding to thermal neutrality, but the one that decreased the sensation of warmth without causing too much discomfort due to draft. Compared with the neutral one along with large nonuniformity, subjects perceive the slightly warm environment with small non-uniformity more comfortable because they evaluate thermal environment based on their perceptions of overall thermal sensation and non-uniformity of thermal sensation. 4.2. Thermal acceptability and comfort The present study shows that overall thermal acceptability and comfort are correlated closely and acceptable range contains all comfortable and just uncomfortable votes on thermal comfort scale. In the present study overall thermal acceptability and comfort were reported by the subjects at the same time and the simultaneous voting could have some contributions to the strong relationship between the two responses. Another study on heated/ cooled seat performed by the author [6] used alternating voting method which allows thermal acceptability votes to be more independent of thermal comfort votes than would be the case had they been obtained simultaneously and the result is shown in Fig. 11. In this study the same thermal acceptability scale was used (see Fig. 1b) and the 4-point thermal comfort scale (+1: Comfortable, +2: Slightly uncomfortable, +3: Uncomfortable, +4: Very uncomfortable) was adopted instead of the scale used in the present study. It can be seen from the figure that the two responses are highly correlated ðR2 ¼ 0:99Þ and both comfortable and slightly uncomfortable votes are acceptable to the subjects, which confirms the results of the present study. There is still one difference between the results of the two studies. ‘Comfort’ corresponds to thermal acceptability mean vote of 0.5 in the present study but to 1 in the study on heated/cooled seat. This can be explained as the different
Overall thermal acceptability
Y. Zhang, R. Zhao / Building and Environment 43 (2008) 44–50
49
1.0 TA = -0.69*TC + 1.62 R2 = 0.99, p < 0.001
0.5 0.0 -0.5 -1.0 1.0
2.0 3.0 Overall thermal comfort
4.0
Fig. 11. Overall thermal acceptability as a function of overall thermal comfort (obtained in the study on heated/cooled seat [6]).
thermal comfort scales used in the two studies. The new thermal comfort scale used in the present study covers wider range than the 4-point scale and people seem likely to perceive the most comfortable category, which is ‘Very comfortable’, instead of ‘Comfortable’ ’as ‘Clearly acceptable’ when they are using the new thermal comfort scale. 4.3. Applications and further studies The effects of various non-uniformities of thermal environment on thermal acceptability were studied previously by many researchers [24]. In those studies the subjects were always in thermal neutrality and exposed only to the non-uniformities of the environment. Recently the studies are transferred to focus on the effect of local cooling or heating on thermal acceptability when the whole body is warm or cool. The results obtained in thermal neutrality cannot be applied directly to the warm or cool conditions. The present study extends the range of the test conditions and derives the relationship applicable to a broad range of whole body thermal state from neutral to warm. Though obtained under the conditions of local cooling, nonuniformity of thermal sensation distribution over the entire body may be an important factor to determine human’s thermal acceptability with various non-uniform environments, such as asymmetric thermal radiation, vertical air temperature difference and warm and cold floors and so on. The relationship proposed by Fanger is validated under uniform conditions and the acceptable range can be determined by the response of whole body thermal sensation. However, it would be inadequant to evaluate overall thermal acceptability only by overall thermal sensation under non-uniform conditions. Non-uniformity of thermal sensation is another key factor and the new thermal acceptability model which combines the effects of overall thermal sensation and non-uniformity of thermal sensation would be very useful for the further chamber and field studies and practical applications relevant to thermally non-uniform environment. The breakage of the relationship between thermal sensation and comfort is observed experimentally as well under dynamic thermal environment by many researchers [4,25,26]. Zhao [27] made a systemic comparison of thermal
ARTICLE IN PRESS 50
Y. Zhang, R. Zhao / Building and Environment 43 (2008) 44–50
comfort under steady and transient conditions and concluded that the meaning of thermal comfort changed with the conditions tested. Thermal comfort means the state of no difference under steady conditions and pleasure associated with the relief of thermal discomfort under transient conditions. Apparently the results obtained under steady state can not be applied directly to non-steady state and the relationship between thermal sensation, acceptability and comfort under dynamic or transient conditions needs further studies.
[3] [4]
[5]
[6]
5. Conclusions The relationships between overall thermal sensation, acceptability and comfort under uniform and non-uniform conditions were studied in the present study and the following conclusions can be drawn: 1. Overall thermal acceptability and comfort are highly correlated and the thermal comfort votes falling in comfortable or slightly uncomfortable range are perceived as acceptable under uniform and non-uniform conditions. 2. Overall thermal sensation is correlated with thermal acceptability and comfort closely and acceptable range runs from neutral to 1.5 (midpoint between ‘Slightly Warm’ and ‘Warm’) on thermal sensation scale under uniform conditions. However, overall thermal sensation is apart from the other two responses and thermal neutrality does not necessarily represent thermal comfort under non-uniform conditions. 3. Non-uniformity of thermal sensation, which is an important factor to determine thermal acceptability with non-uniform environment, well explains the breakage of the relationship between overall thermal sensation and acceptability and comfort. 4. Combining the effects of overall thermal sensation and non-uniformity of thermal sensation, a new thermal acceptability model is proposed. The model is applicable to uniform and non-uniform conditions over a wide range of whole body thermal state from neutral to warm.
Acknowledgements
[7] [8]
[9]
[10]
[11]
[12]
[13] [14]
[15] [16]
[17]
[18] [19] [20]
[21] [22]
The experiment was supported by National Natural Science Foundation of China (Project No. 50178037) and National Natural Science Foundation of China (Project No. 50538040). References [1] Burch SD, Pearson JT, Ramadhyani S. Experimental study of passenger thermal comfort in an automobile under severe winter conditions. ASHRAE Transactions 1991;97(1):239–46. [2] Hagino M, Hara J. Development of a method for predicting comfortable airflow in the passenger compartment. SAE technical
[23]
[24] [25] [26]
[27]
paper series no. 922131. New York: Society of Automotive Engineers; 1992. Li J. Studies on characteristics of personalized ventilation and human response. PhD thesis, Tsinghua University, Beijing, 2004 [in Chinese]. Hodder SG, Loveday DL, Parsons KC, et al. Thermal comfort in chilled ceiling and displacement ventilation environments: vertical radiant temperature asymmetry effects. Energy and Buildings 1998;27:167–73. Melikov AK, Arakelian RS, Halkjaer L, et al. Spot cooling, part 1: human responses to cooling with air jets. ASHRAE Transactions 1994;100(2):476–99. Zhang Y, Wyon DP, Fang L, et al. Optimal seat control algorithm. Research project final report, Johnson Controls Inc; 2003 [unpublished]. Melikov AK. Personalized ventilation. Indoor Air 2004;14(7):157–67. Gong N, Tham KW, Melikov AK. Human perception of local air movement and the acceptable air velocity range for local air movement in the tropics. In: Proceedings of indoor air 2005, Beijing, China 2005. p. 452–7. Schlegel JC, McNall PE. The effect of asymmetric radiation on the thermal and comfort sensations of sedentary subjects. ASHRAE Transactions 1968;74(2):144–54. Brooks JE, Parsons KC. An ergonomics investigation into human thermal comfort using an automobile seat heated with encapsulated carbonized fabric. Ergonomics 1999;42(5):661–73. Zhang H. Human thermal sensation and comfort in transient and non-uniform thermal environments. PhD thesis, University of California at Berkeley, Berkeley; 2003. Sakoi T, Tsuzuki K, Kato S, et al. Proposal of expression of wholebody thermal comfort in non-uniform thermal environment. Proceedings of indoor air 2005, Beijing, China, 2005. p. 193–8. Fanger PO. Thermal comfort. Copenhagen: Danish Technical Press; 1970. Gagge AP, Stolwijk JAJ, Hardy JD. Comfort and thermal sensations and associated physiological responses at various ambient temperatures. Environmental Research 1967;1:1–20. Berglund LG. Thermal acceptability. ASHRAE Transactions 1979;85(2):825–34. Gagge AP, Nevins RG. Effect of energy conservation guideline, on comfort, acceptability and health. Final report of contract #CO-0451891-00, Federal Energy Administration, USA; 1976. Berglund LG, Gonzalez RR. Application of acceptable temperature drifts to built environments as a mode of energy conservation. ASHRAE Transactions 1978;84(1):110–21. Gunnarsen L, Fanger PO. Adaptation to indoor air pollution. Environment International 1992;18:43–54. Wyon DP. Methodology for indoor environmental research. ICIE symposium; 2001. McNall PE, Biddison RE. Thermal and comfort sensations of sedentary persons exposed to asymmetric radiant fields. ASHRAE Transactions 1970;76(1):123–36. Hensel H. Thermoreception and temperature regulation. London: Academic Press; 1981. McIntyre DA. Indoor climate. London: Applied Science Publishers Ltd; 1980. Olesen BW and Nielsen R. Convective spot cooling of hot working environment. In: 16th international conference of refrigeration, Commission E1, Paris, 1983. p. 201–7. ASHRAE. Handbook of fundamentals. Atlanta: American Society of Heating, Refrigerating and Air conditioning Engineers, Inc.; 2001. Chatonnet J, Cabanac M. The perception of thermal comfort. International Journal of Biometeorology 1965;9:183–93. Mower GD. Perceived intensity of peripheral thermal stimuli is independent of internal body temperature. Journal of Comparative & Physiological Psychology 1976;90:1152–5. Zhao RY. Discussion on thermal comfort. Heating Ventilation and Air Conditioning 2000;30(3):25–6.
Building and Environment 43 (2008) 44–50 www.elsevier.com/locate/buildenv
Overall thermal sensation, acceptability and comfort Yufeng Zhanga,, Rongyi Zhaob a
Department of Architecture, South China University of Technology, Wushan, Guangzhou 510640, PR China b Department of Building Science, Tsinghua University, Beijing 100084, PR China Received 6 July 2006; received in revised form 8 November 2006; accepted 27 November 2006
Abstract The relationships between overall thermal sensation, acceptability and comfort were studied experimentally under uniform and nonuniform conditions separately. Thirty subjects participated in the experiment and reported their local thermal sensation of each body part, overall thermal sensation, acceptability and comfort simultaneously. Sensation, acceptability and comfort were found to be correlated closely under uniform conditions and acceptable range ran from neutral to 1.5 (midpoint between ‘Slightly Warm’ and ‘Warm’) on thermal sensation scale and contained all comfortable and slightly uncomfortable votes on thermal comfort scale. Under non-uniform conditions overall thermal acceptability and comfort were correlated closely. However, overall thermal sensation was apart from the other two responses and non-uniformity of thermal sensation was found to be the reason for the breakage. Combining the effects of overall thermal sensation and non-uniformity of thermal sensation, a new thermal acceptability model was proposed and the model was testified to be applicable to uniform and non-uniform conditions over a wide range of whole body thermal state from neutral to warm. r 2007 Elsevier Ltd. All rights reserved. Keywords: Thermal sensation; Thermal acceptability; Thermal comfort; Non-uniform environment; Non-uniformity of thermal sensation
1. Introduction With the requirements of energy saving, more and more attention is paid recently to the study on thermally nonuniform environment. Assessment of non-uniform environment is a highlighted problem for well-designed nonuniform environment practice in buildings. For the assessment of uniform and steady thermal environment, the known indices, such as PMV, ET* and SET, which predict human thermal sensation by environmental parameters and personal informations, are widely accepted. However, there is no universal index to evaluate thermally non-uniform environment, and overall (whole body) thermal sensation [1–4], overall thermal acceptability [5–8] and overall thermal comfort [9–12] were used separately by different researchers. It would be useful to understand the relationship between overall thermal sensation, acceptability and comfort under non-uniform environment. Corresponding author. Tel.: +86 20 39851485; fax: +86 20 87110164.
E-mail address: [email protected] (Y. Zhang). 0360-1323/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.buildenv.2006.11.036
The relationship between thermal sensation and acceptability was firstly clarified by Fanger [13], who defined the dissatisfied based on the experimental results by Gagge et al. [14] as those who vote ‘Cool’ or ‘Cold’, ‘Warm’ or ‘Hot’. This definition was confirmed by Berglund [15] through comparison with the responses obtained by directly asking subjects whether they find the thermal conditions acceptable or unacceptable. As the relationship was derived and confirmed under uniform conditions, its validity under non-uniform conditions remains untested. The relationship between thermal comfort and acceptability was investigated by Berglund [15]. He compared the effect of temperatures that deviate from those of optimum comfort assessed by percent comfortable [16] with the one by thermal acceptability [17] and found that they were quite similar, which indicates that the thermal comfort votes falling in comfortable or slightly uncomfortable range were perceived by the subjects as acceptable. The comparison was conducted under uniform environment and the one for non-uniform environment remains vacant. The known relationships between thermal sensation, acceptability and comfort were derived qualitatively under
ARTICLE IN PRESS Y. Zhang, R. Zhao / Building and Environment 43 (2008) 44–50
uniform environment. The purpose of the present study was to investigate quantitatively the relationships between overall thermal sensation, acceptability and comfort under uniform and non-uniform environments and to develop a new thermal acceptability model applicable to both environments. 2. Experimental methods The experiment was carried out in the climate chamber in Tsinghua University. The chamber was used to achieve a thermally uniform environment and the ambient air temperature in the chamber was maintained with a precision of 0:2 C. Personalized ventilation system was used to produce non-uniform environment by supplying local cooling airflow to three sensitive segments of human body—face, chest and back—separately. Three room temperatures, ranging from neutral to warm, were chosen and for each room temperature, three local cooling target temperatures (target temperature means the air temperature at the center of cooling body part surface), ranging from neutral to slightly cool, were studied in the present study (Table 1). Relative humidity was kept at 40%, and air velocity was less than 0.1 m/s in the room air. Air velocity at the outlet of local cooling airflow was maintained at 1 m/s. Each test consisted of half-an-hour exposure to uniform condition and half-an-hour exposure to non-uniform condition. The ambient room temperature was maintained Table 1 Experimental conditions Room temperature (1C) Target temperature (1C)
28–32–35 22–25–28
a Hot
+2
Warm
+1
Slightly warm
+1
0
the same and local cooling airflow was supplied only when the exposure to non-uniform condition started. Thirty randomly selected Chinese students, with a normal range of age, height and weight, participated in the experiment of all conditions and the total duration for each subject was 27 h. All the subjects wore only shorts during the experiment to keep the same clothing insulation for the three cooling body segments. The sequence of presentation was balanced among the subjects. Three subjects participated in the test at one time and remained sedentary throughout each exposure. Conversation was permitted but the subjects were not allowed to exchange views concerning the thermal environment. Subjects reported their local thermal sensation of each body part, overall thermal sensation, overall thermal acceptability and overall thermal comfort simultaneously at each voting time and three times in the last 10 min for each exposure. Thermal sensations were reported on ASHRAE 7-point scale (Fig. 1a). A visual-analogue scale indicating acceptability, originally developed to evaluate indoor air quality by Gunnarsen and Fanger [18], was used in the present study (Fig. 1b). A thermal comfort scale developed by Zhang [11] was applied in the present study to force subjects to make a clear determination about whether their perceived state falls in the category of ‘Comfortable’ or ‘Uncomfortable’ (Fig. 1c). 3. Results and discussion Shapiro–Wilk’s W test was applied and the results show that human responses obtained in all conditions were normally distributed. They were therefore analyzed using repeated-measure ANOVA. Compared with independentmeasure design, repeated-measure design with balanced order of presentation is a more efficient approach, which
b
+3
Neutral
-1
Slightly cool
-2
Cool
-3
Cold
0 0
c +2
Very comfortable
+1
Comfortable
Clearly acceptable
Just acceptable
0
Just comfortable
0
Just uncomfortable
Just unacceptable -1
Thermal sensation scale
-1
45
Clearly unacceptable -2
Thermal acceptability scale
Uncomfortable
Very uncomfortable Thermal comfort scale
Fig. 1. Voting scales: (a) thermal sensation scale; (b) thermal acceptability scale and (c) thermal comfort scale.
ARTICLE IN PRESS Y. Zhang, R. Zhao / Building and Environment 43 (2008) 44–50
1.2
0.8 Overall thermal comfort
Overall thermal acceptability
46
0.4
0.0 TA = -0.41*TS + 0.58 R2 = 0.96, p < 0.001 -0.4 -0.5
0.0 0.5 1.0 1.5 Overall thermal sensation
2.0
0.6
0.0 TC = -0.79*TS + 1.01 R2 = 0.96, p < 0.001 -0.6 -0.5
0.0 0.5 1.0 1.5 Overall thermal sensation
2.0
means the same level of accuracy can be achieved by using fewer subjects. Thirty subjects in repeated-measure design with balanced order of presentation, which was used successfully in the study of environmental effects on simulated office work in a closely controlled environment [19], was applied in the present study to attain the test accuracy efficiently. It was found that human responses reached steady state within 20 min during each exposure ðp40:05Þ. If not mentioned specifically, all responses reported below are steady state responses. 3.1. Uniform environment
Overall thermal acceptability
Fig. 2. Overall thermal acceptability, TA, and overall thermal comfort, TC, as functions of overall thermal sensation, TS, under the uniform conditions.
0.8
0.4
0.0
-0.4 -0.4
TA = 0.51*TC + 0.07 R2 = 0.99, p< 0.001 0.0 0.4 0.8 Overall thermal comfort
1.2
Fig. 3. Overall thermal acceptability as a function of overall thermal comfort under the uniform conditions.
Human responses to the uniform conditions were analyzed and the results are shown in Figs. 2 and 3. Each point in the figures represents the mean vote of all subjects’ responses for each condition. It can be seen that a straight line fits the data well for each figure ðR2 40:96Þ. Under the uniform conditions, overall thermal sensation, acceptability and comfort are correlated with each other closely. Thermal sensation mean vote of 1.5 corresponds to thermal acceptability mean vote of 0 and thermal comfort mean vote of 0:2, that is to say, acceptable range runs from neutral to 1.5 on thermal sensation scale and contains all comfortable and slightly uncomfortable votes on thermal comfort scale, which confirms the relationships found by others [13,15] under uniform conditions.
similar with the one obtained under the uniform conditions. The scattered points in Fig. 4 indicate that it is difficult to define acceptable or comfortable range only by whole body thermal sensation vote as each thermal sensation vote corresponds to a wide range of thermal acceptability and comfort. For instance, thermal neutrality corresponds to the acceptability range of (0.07, 0.54) and the comfort range of ð0:02; 0:97Þ. Thus the relationship proposed by Fanger is not suitable for non-uniform environment. Except for overall thermal sensation, there should exist other important factors influencing thermal acceptability and comfort under non-uniform conditions.
3.2. Non-uniform environment
3.3. Non-uniformity of thermal sensation
Human responses to the non-uniform conditions were analyzed and the results are shown in Figs. 4 and 5. Each point represents the mean vote of all subjects’ responses for each condition. Compared with the linear relationships obtained in the uniform conditions, no linear relationships exist between overall thermal sensation and acceptability and comfort under the non-uniform conditions (see Fig. 4). Subjects feel more uncomfortable and unacceptable with the non-uniform environment than the uniform one while their overall thermal sensations maintain the same. The linear relationship between overall thermal acceptability and comfort is retained well ðR2 ¼ 0:98Þ under the nonuniform conditions (see Fig. 5) and the function is very
McNall and Biddison [20] studied thermal sensation and comfort of sedentary persons exposed to asymmetric radiant fields and found that it was ‘uneven body temperature’ which caused the thermally neutral subjects participating in the Hot Wall series to have a significantly lower probability of feeling comfortable than the subjects in the uniform conditions, where the ‘uneven body temperature’ means one side of the body feels warmer (or cooler) than the other. ‘Uneven body temperature’ was investigated by using an additional questionnaire in the present study and it was found that 97% of the subjects perceived obvious non-uniformity of thermal sensation between different body parts during the non-uniform
ARTICLE IN PRESS 0.8 Overall thermal comfort
Overall thermal acceptability
Y. Zhang, R. Zhao / Building and Environment 43 (2008) 44–50
0.4
0.0
47
1.2 0.8 0.4 0.0 -0.4
-0.4 -1
1 2 0 Overall thermal sensation Non-uniform
-1
0 1 Overall thermal sensation Uniform
2
0.8 0.4 0.0 TA = 0.52*TC + 0.06 R2 = 0.98, p < 0.001 -0.4 -0.4
0.0 0.4 0.8 Overall thermal comfort
1.2
Overall thermal acceptability
Overall thermal acceptability
Fig. 4. Overall thermal acceptability and comfort as functions of overall thermal sensation under the non-uniform conditions.
0.8 TA = -0.27*TSD + 0.57 R2 = 0.89, p < 0.001
0.6 0.4 0.2 0 -0.2 0.0
Fig. 5. Overall thermal acceptability as a function of overall thermal comfort under the non-uniform conditions.
3.4. A new thermal acceptability model Subjects evaluate thermally non-uniform environment based on their perceptions of overall thermal sensation and non-uniformity of thermal sensation. Supposing these two perceptions are independent from each other, a new
2.0
Fig. 6. Overall thermal acceptability as a function of the maximum thermal sensation difference between body parts, TSD , when in overall thermal neutrality.
1.4 Overall thermal comfort
exposures. Non-uniformity of thermal sensation may be the reason for the scattering of the points in Fig. 4. Considering the strongest feeling comes from the difference between the coolest and the warmest body parts, the maximum thermal sensation difference between body parts was chosen to represent the non-uniformity of thermal sensation. Taking the responses obtained when the subject’s whole body thermal sensation votes close to neutral (mean vote of overall thermal sensation falls in ð0:2; þ0:2Þ), the relationships between the maximum thermal sensation difference between body parts and overall thermal comfort and acceptability were analyzed and the results are shown in Figs. 6 and 7. When the nonuniformity is small, subjects perceive the non-uniform environment to be similarly comfortable and acceptable as the uniform one. With the increase of non-uniformity, overall thermal comfort and acceptability go down apparently and linearly ðR2 40:88Þ. Non-uniformity of thermal sensation well explains the breakage of the relationship between overall thermal sensation and acceptability and comfort.
0.4 0.8 1.2 1.6 The maximum thermal sensation difference between body parts
TC = -0.55*TSD + 1.01 R2 = 0.88, p< 0.001
1 0.6 0.2 -0.2 0.0
0.4 0.8 1.2 1.6 The maximum thermal sensation difference between body parts
2.0
Fig. 7. Overall thermal comfort as a function of the maximum thermal sensation difference between body parts when in overall thermal neutrality.
thermal acceptability model was proposed: TA ¼ TA1 þ TA2 ,
(1)
where TA is overall thermal acceptability, TA1 is the uniform term and TA2 is the non-uniform term. The uniform term is a function of overall thermal sensation TS, and the function was obtained by linear regression of the data obtained under the uniform
ARTICLE IN PRESS Y. Zhang, R. Zhao / Building and Environment 43 (2008) 44–50 Predicted thermal acceptability
48
way under uniform environment and non-uniform environment. The effect of non-uniformity of thermal sensation on thermal acceptability obtained under thermal neutral conditions is valid for warm conditions. The acceptable range according to the new thermal acceptability model can be expressed as
0.6 TS (0,0.5) TS (0.5,1.5) 0.3
0.0 R2 = 0.91 p < 0.001 -0.3 -0.3
0.0 0.3 Measured thermal acceptability
0:41 TS þ 0:58 0:27TSD 40.
0.6
4. Discussion
Predicted thermal acceptability
Fig. 8. Validation of the new thermal acceptability model in the nonuniform conditions that vary from neutrality.
0.6
0.3
0 R2 = 0.98 p < 0.001 -0.3 -0.3
0 0.3 Measured thermal acceptability
0.6
Fig. 9. Validation of the new thermal acceptability model in the uniform conditions.
conditions (see Fig. 2): TA1 ¼ 0:41 TS þ 0:58.
4.1. Thermal sensation and comfort The present study indicates that the mean vote obtained by thermal sensation scale is in good agreement with the one by thermal comfort scale and thus it is reasonable to use only one of the two scales or combine them together into a single scale to assess the thermally uniform environment. A good example for the latter one is the well-known Bedford scale (see Fig. 10), which confounds thermal sensation and comfort and employs the terms ‘too’ and ‘much too’ indicating that the observer is not satisfied with his thermal environment [21]. McIntyre [22] indicated that the results obtained by the Bedford scale and the ASHRAE 7-point scale may be compared directly with each other, which confirms the findings obtained in the present study. However, the close relationship between thermal sensation and comfort does not exist under the +3
Much too warm
+2
Too warm
+1
Comfortable warm
(2)
The non-uniform term is a function of the maximum thermal sensation difference between body parts TSD . The function was obtained by linear regression of the data under the non-uniform conditions while whole body thermal sensation closes to neutral (see Fig. 6): TA2 ¼ 0:27 TSD .
(4)
(3)
In order to validate the new thermal acceptability model in the non-uniform conditions that vary from neutrality and the uniform conditions, all responses obtained in the present study were collected and the predicted overall thermal acceptability calculated from the actual overall thermal sensation and non-uniformity of thermal sensation was compared with the actual overall thermal acceptability (see Figs. 8 and 9). The responses in neutral–cool range were not included in the analysis as Eq. (2) was obtained in neutral–warm conditions. It can be seen that the predicted values are highly correlated with the measured values ðR2 40:91; po0:001Þ. The new thermal acceptability model is suitable for both neutral and warm, uniform and non-uniform conditions. Subjects evaluate the acceptability of thermal environment from two aspects separately: whole body thermal sensation and non-uniformity of thermal sensation. Whole body thermal sensation affects thermal acceptability in the same
0
Comfortable
-1
Comfortable cool
-2
Too cool
-3
Much too cool
Fig. 10. Bedford scale.
ARTICLE IN PRESS non-uniform conditions as shown in the present study, which indicates that the Bedford scale is not applicable to the assessment of thermal comfort in non-uniform environment. The best way is to directly ask subjects whether or not they feel comfortable or to report their perceptions on the thermal comfort scale. Thermal neutrality for a person is defined as the condition in which the subject would prefer neither warmer nor cooler surroundings. Usually thermal neutrality is regarded the same as thermal comfort, which is confirmed by the present study (see Fig. 2) under the uniform conditions. Under the non-uniform conditions, thermal neutrality unnecessarily correspond to thermal comfort. The more the non-uniformity of thermal sensation perceived by the subjects, the more the discomfort they report even when they are in overall thermal neutrality (see Fig. 7). This is confirmed by the experimental results obtained by Olesen and Nielsen [23] and Melikov et al. [5]. They studied human responses to local cooling with air jets in warm conditions and found that the air jet velocity preferred by the subjects was not the one corresponding to thermal neutrality, but the one that decreased the sensation of warmth without causing too much discomfort due to draft. Compared with the neutral one along with large nonuniformity, subjects perceive the slightly warm environment with small non-uniformity more comfortable because they evaluate thermal environment based on their perceptions of overall thermal sensation and non-uniformity of thermal sensation. 4.2. Thermal acceptability and comfort The present study shows that overall thermal acceptability and comfort are correlated closely and acceptable range contains all comfortable and just uncomfortable votes on thermal comfort scale. In the present study overall thermal acceptability and comfort were reported by the subjects at the same time and the simultaneous voting could have some contributions to the strong relationship between the two responses. Another study on heated/ cooled seat performed by the author [6] used alternating voting method which allows thermal acceptability votes to be more independent of thermal comfort votes than would be the case had they been obtained simultaneously and the result is shown in Fig. 11. In this study the same thermal acceptability scale was used (see Fig. 1b) and the 4-point thermal comfort scale (+1: Comfortable, +2: Slightly uncomfortable, +3: Uncomfortable, +4: Very uncomfortable) was adopted instead of the scale used in the present study. It can be seen from the figure that the two responses are highly correlated ðR2 ¼ 0:99Þ and both comfortable and slightly uncomfortable votes are acceptable to the subjects, which confirms the results of the present study. There is still one difference between the results of the two studies. ‘Comfort’ corresponds to thermal acceptability mean vote of 0.5 in the present study but to 1 in the study on heated/cooled seat. This can be explained as the different
Overall thermal acceptability
Y. Zhang, R. Zhao / Building and Environment 43 (2008) 44–50
49
1.0 TA = -0.69*TC + 1.62 R2 = 0.99, p < 0.001
0.5 0.0 -0.5 -1.0 1.0
2.0 3.0 Overall thermal comfort
4.0
Fig. 11. Overall thermal acceptability as a function of overall thermal comfort (obtained in the study on heated/cooled seat [6]).
thermal comfort scales used in the two studies. The new thermal comfort scale used in the present study covers wider range than the 4-point scale and people seem likely to perceive the most comfortable category, which is ‘Very comfortable’, instead of ‘Comfortable’ ’as ‘Clearly acceptable’ when they are using the new thermal comfort scale. 4.3. Applications and further studies The effects of various non-uniformities of thermal environment on thermal acceptability were studied previously by many researchers [24]. In those studies the subjects were always in thermal neutrality and exposed only to the non-uniformities of the environment. Recently the studies are transferred to focus on the effect of local cooling or heating on thermal acceptability when the whole body is warm or cool. The results obtained in thermal neutrality cannot be applied directly to the warm or cool conditions. The present study extends the range of the test conditions and derives the relationship applicable to a broad range of whole body thermal state from neutral to warm. Though obtained under the conditions of local cooling, nonuniformity of thermal sensation distribution over the entire body may be an important factor to determine human’s thermal acceptability with various non-uniform environments, such as asymmetric thermal radiation, vertical air temperature difference and warm and cold floors and so on. The relationship proposed by Fanger is validated under uniform conditions and the acceptable range can be determined by the response of whole body thermal sensation. However, it would be inadequant to evaluate overall thermal acceptability only by overall thermal sensation under non-uniform conditions. Non-uniformity of thermal sensation is another key factor and the new thermal acceptability model which combines the effects of overall thermal sensation and non-uniformity of thermal sensation would be very useful for the further chamber and field studies and practical applications relevant to thermally non-uniform environment. The breakage of the relationship between thermal sensation and comfort is observed experimentally as well under dynamic thermal environment by many researchers [4,25,26]. Zhao [27] made a systemic comparison of thermal
ARTICLE IN PRESS 50
Y. Zhang, R. Zhao / Building and Environment 43 (2008) 44–50
comfort under steady and transient conditions and concluded that the meaning of thermal comfort changed with the conditions tested. Thermal comfort means the state of no difference under steady conditions and pleasure associated with the relief of thermal discomfort under transient conditions. Apparently the results obtained under steady state can not be applied directly to non-steady state and the relationship between thermal sensation, acceptability and comfort under dynamic or transient conditions needs further studies.
[3] [4]
[5]
[6]
5. Conclusions The relationships between overall thermal sensation, acceptability and comfort under uniform and non-uniform conditions were studied in the present study and the following conclusions can be drawn: 1. Overall thermal acceptability and comfort are highly correlated and the thermal comfort votes falling in comfortable or slightly uncomfortable range are perceived as acceptable under uniform and non-uniform conditions. 2. Overall thermal sensation is correlated with thermal acceptability and comfort closely and acceptable range runs from neutral to 1.5 (midpoint between ‘Slightly Warm’ and ‘Warm’) on thermal sensation scale under uniform conditions. However, overall thermal sensation is apart from the other two responses and thermal neutrality does not necessarily represent thermal comfort under non-uniform conditions. 3. Non-uniformity of thermal sensation, which is an important factor to determine thermal acceptability with non-uniform environment, well explains the breakage of the relationship between overall thermal sensation and acceptability and comfort. 4. Combining the effects of overall thermal sensation and non-uniformity of thermal sensation, a new thermal acceptability model is proposed. The model is applicable to uniform and non-uniform conditions over a wide range of whole body thermal state from neutral to warm.
Acknowledgements
[7] [8]
[9]
[10]
[11]
[12]
[13] [14]
[15] [16]
[17]
[18] [19] [20]
[21] [22]
The experiment was supported by National Natural Science Foundation of China (Project No. 50178037) and National Natural Science Foundation of China (Project No. 50538040). References [1] Burch SD, Pearson JT, Ramadhyani S. Experimental study of passenger thermal comfort in an automobile under severe winter conditions. ASHRAE Transactions 1991;97(1):239–46. [2] Hagino M, Hara J. Development of a method for predicting comfortable airflow in the passenger compartment. SAE technical
[23]
[24] [25] [26]
[27]
paper series no. 922131. New York: Society of Automotive Engineers; 1992. Li J. Studies on characteristics of personalized ventilation and human response. PhD thesis, Tsinghua University, Beijing, 2004 [in Chinese]. Hodder SG, Loveday DL, Parsons KC, et al. Thermal comfort in chilled ceiling and displacement ventilation environments: vertical radiant temperature asymmetry effects. Energy and Buildings 1998;27:167–73. Melikov AK, Arakelian RS, Halkjaer L, et al. Spot cooling, part 1: human responses to cooling with air jets. ASHRAE Transactions 1994;100(2):476–99. Zhang Y, Wyon DP, Fang L, et al. Optimal seat control algorithm. Research project final report, Johnson Controls Inc; 2003 [unpublished]. Melikov AK. Personalized ventilation. Indoor Air 2004;14(7):157–67. Gong N, Tham KW, Melikov AK. Human perception of local air movement and the acceptable air velocity range for local air movement in the tropics. In: Proceedings of indoor air 2005, Beijing, China 2005. p. 452–7. Schlegel JC, McNall PE. The effect of asymmetric radiation on the thermal and comfort sensations of sedentary subjects. ASHRAE Transactions 1968;74(2):144–54. Brooks JE, Parsons KC. An ergonomics investigation into human thermal comfort using an automobile seat heated with encapsulated carbonized fabric. Ergonomics 1999;42(5):661–73. Zhang H. Human thermal sensation and comfort in transient and non-uniform thermal environments. PhD thesis, University of California at Berkeley, Berkeley; 2003. Sakoi T, Tsuzuki K, Kato S, et al. Proposal of expression of wholebody thermal comfort in non-uniform thermal environment. Proceedings of indoor air 2005, Beijing, China, 2005. p. 193–8. Fanger PO. Thermal comfort. Copenhagen: Danish Technical Press; 1970. Gagge AP, Stolwijk JAJ, Hardy JD. Comfort and thermal sensations and associated physiological responses at various ambient temperatures. Environmental Research 1967;1:1–20. Berglund LG. Thermal acceptability. ASHRAE Transactions 1979;85(2):825–34. Gagge AP, Nevins RG. Effect of energy conservation guideline, on comfort, acceptability and health. Final report of contract #CO-0451891-00, Federal Energy Administration, USA; 1976. Berglund LG, Gonzalez RR. Application of acceptable temperature drifts to built environments as a mode of energy conservation. ASHRAE Transactions 1978;84(1):110–21. Gunnarsen L, Fanger PO. Adaptation to indoor air pollution. Environment International 1992;18:43–54. Wyon DP. Methodology for indoor environmental research. ICIE symposium; 2001. McNall PE, Biddison RE. Thermal and comfort sensations of sedentary persons exposed to asymmetric radiant fields. ASHRAE Transactions 1970;76(1):123–36. Hensel H. Thermoreception and temperature regulation. London: Academic Press; 1981. McIntyre DA. Indoor climate. London: Applied Science Publishers Ltd; 1980. Olesen BW and Nielsen R. Convective spot cooling of hot working environment. In: 16th international conference of refrigeration, Commission E1, Paris, 1983. p. 201–7. ASHRAE. Handbook of fundamentals. Atlanta: American Society of Heating, Refrigerating and Air conditioning Engineers, Inc.; 2001. Chatonnet J, Cabanac M. The perception of thermal comfort. International Journal of Biometeorology 1965;9:183–93. Mower GD. Perceived intensity of peripheral thermal stimuli is independent of internal body temperature. Journal of Comparative & Physiological Psychology 1976;90:1152–5. Zhao RY. Discussion on thermal comfort. Heating Ventilation and Air Conditioning 2000;30(3):25–6.