Physiological and subjective thermal response from Indians

Building and Environment 70 (2013) 306e317

Contents lists available at ScienceDirect

Building and Environment journal homepage: www.elsevier.com/locate/buildenv

Physiological and subjective thermal response from Indians Rina Maiti* Human Engineering Research Laboratory (HERL), Centre for Product Design & Manufacturing (CPDM), Indian Institute of Science (IISc), Bangalore 560012, India

a r t i c l e i n f o

a b s t r a c t

Article history: Received 30 April 2013 Received in revised form 12 August 2013 Accepted 21 August 2013

A controlled laboratory experiment was carried out on forty Indian male college students for evaluating the effect of indoor thermal environment on occupants’ response and thermal comfort. During experiment, indoor temperature varied from 21
C to 33
C, and the variables like relative humidity, airflow, air temperature and radiant temperature were recorded along with skin (Tsk) and oral temperature (Tcore) from the subjects. From Tsk and Tc, body temperature (Tb) was evaluated. Subjective Thermal Sensation Vote (TSV) was recorded using ASHRAE 7-point scale. In PMV model, Fanger’s Tsk equation was used to accommodate adaptive response. Stepwise regression analysis result showed Tb was better predictor of TSV than Tsk and Tcore. Regional skin temperature response, lower sweat threshold temperature with no dipping sweat and higher cutaneous sweating threshold temperature were observed as thermal adaptive responses. Using PMV model, thermal comfort zone was evaluated as (22.46e25.41)
C with neutral temperature of 23.91
C, whereas using TSV response, wider comfort zone was estimated as (23.25 e26.32)
C with neutral temperature at 24.83
C. It was observed that PMV-model overestimated the actual thermal response. Interestingly, these subjects were found to be less sensitive to hot but more sensitive to cold. A new TSV-PPD relation (PPDnew) was obtained with an asymmetric distribution of hotcold thermal sensation response in Indians. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Thermal comfort Thermal Sensation Vote Indian climate PMV and TSV Tropical thermal adaptation

1. Introduction A healthy and comfortable thermal environment of indoor workspace makes the occupants comfortable with improved work efficiencies and well-being. Thermal sensation refers to subjective thermal perception as how the person feels the environment e warm, neutral, cold etc. Different environmental parameters, activity level and clothing worn by the occupants affect thermal sensation experienced by a person. The word “thermal comfort” is defined by American Society of Heating, Refrigerating and AirConditioning Engineers (ASHRAE), as the condition of mind which expresses human satisfaction with the thermal environment [1,2]. To evaluate subjective thermal sensation, International standards such as ISO 7730 (2005) [3] and the ASHRAE Standard 55-92 [2] describes a method to estimate subjective thermal sensation based on Predicted Mean Vote (PMV) model, developed by Fanger [4]. Otherwise, researcher often ask the subject to rate their response using 7-points ASHRAE voting scale that covers a range of cool (3) to warm (þ3) sensation and to discriminate the result * Centre for Product Design & Manufacturing (CPDM), Indian Institute of Science (IISc), Bangalore 560012, Karnataka, India. Tel.: þ91 80 22933134 (off). E-mail addresses: [email protected], [email protected]. 0360-1323/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.buildenv.2013.08.029

from calculated PMV response, it is generally named as Thermal Sensation Vote (TSV) or Actual Mean Vote (AMV). From this subjective response, thermally comfortable condition (all votes between þ1 (slightly cool) to þ1 (slightly warm)) is evaluated. Fanger’s PMV model calculates subjective response to the thermal environment directly from environmental parameters as it assumes that thermal sensation experienced by a person, a passive response which is a function of physiological strain imposed on him by the environment, and it does not clarify how people respond physiologically and subjectively to the thermal environment. Many researchers are beginning to challenge this concept as they claim that the adaptation (psychological, behavioral, cultural and clothing adaptation) is region specific, which influences person’s thermal preference beyond mere passive experience of a body’s thermal balance and highly influenced by the local climate conditions and socio-cultural set-up, so it may be impossible to assign a specific value to thermal comfort standard [5e7]. In ASHRAE RP 884, “adaptive PMV” is mentioned to include occupant’s adaptive response especially in case of occupants in the naturally ventilated buildings [8]. Several authors have reported that especially the people living in the tropical regions, having higher optimum temperature than those in the cold regions due to the adaptation [9e 14]. Hwang et al. have conducted field experiments in 10

R. Maiti / Building and Environment 70 (2013) 306e317

naturally ventilated and 26 air-conditioned campus classrooms in Taiwan, and found that occupants who have acclimated to the hothumid climate can accept warmer thermal environments [15]. India is having tropical climate with large variation in environmental conditions in different regions [16]. Different inhabited environmental condition produces different subject sensations due to level of adaptation in climatic condition, living patterns, eating habits etc. as reported in case of Indians [17]. Sharma and Ali have assessed thermal sensation for Indians and developed “Tropical Summer Index (TSI)” and identified Indian thermal sensations as “slightly cool”, “comfortable” and “slightly warm” at 19e25
C (with optimum at 22
C), 25e30
C (with optimum at 27.5
C) and 30e34
C (with optimum at 32
C) respectively [17]. Our neighboring country Bangladesh is also having similar type of environmental conditions. Air temperature range of 24e32
C with relative humidity ranging between 50 and 90% and without or in little air movement is reported as thermal comfort zone for Bangladeshi people [18]. Indraganthi has estimated the comfort band for Indians to be (26e32.45)
C with the neutral temperature at 29.23
C [19] from field study. However, the National Building Code of India (BIS, 2005) advocates the use of two indoor temperature ranges for summer (23e26)
C and winter (21e 23)
C for all the climatic zones, which is similar to ASHRAE recommendation but far above than the earlier reported results [20]. For better prediction of thermal response in Indians, it is required to quantify human perception of different thermal conditions and how it is correlated with the physiological parameters. Actually, thermal response is initiated by behavioral thermoregulation, where both mean skin-surface (Tsk) and core (Tcore) temperatures are known afferent inputs to the thermoregulatory system, and contribute about equally toward determining thermal comfort [21]. Hence, both skin and core temperatures are potential physiological parameters for objective evaluation of human thermal sensation experienced by the surrounding environment. Skin temperature initiates thermoregulatory response before activating stronger autonomic and metabolic responses controlled by core temperature to maintain the body temperature. However, estimation of Tsk is relatively easy, while measuring Tcore is always challenging. Rectal temperature is regarded as the most valid core temperature index but it is impractical, invasive and expensive for everyday clinical use. Therefore, oral temperature (recording from sublingual pocket) is one of the common index for Tcore as it’s measurements are easy and reliable. No literature is available on thermal response studies in Indian climate on their natives, and researchers have concluded for the requirement of further study. In this context, a laboratory study is conducted to investigate how mean skin temperature, oral temperature and body temperature are correlated with TSV responses at different thermal environments and any adaptive variations in objective measurement. A new relation is aimed to establish on TSV and the number of responder at different air temperatures, similar to Predicted Percentage of Dissatisfaction (PPD) in PMV model to highlight the thermal adaptive response of Indians. This study also aims to evaluate indoor thermal comfort condition for Indians from this TSV-PPD response, and compare the result with PMV-PPD model estimated result. 2. Methodology 2.1. Subjects Forty young male university students having age e (25.18  2.4) years, weight e (68.6  8.46) kg and height e (1.71  0.05) m participated in this study as volunteers. They were originally from different part of the country and none of them were professional athletes. The Body Mass Index (BMI) was calculated as ratio of body

307

weight (kg) and square of body height (m) and obtained as 23.38  2.03. Body Surface Area (BSA) of individual participant was calculated using DuBois and DuBois [22] equation, used to normalize the heat flow parameters. Prior to the participation, subjects were informed in details about the experiment and their consents were collected. To have uniform effect of local climatic acclimatization, no subject was selected as a new comer to this environment. All these subjects spent at least one year in this university campus. Subjects were instructed not to consume alcohol or medical drugs within 48 h before testing and not to have any food (except water) for at least two hours before the experiment and no water intake during the experiment. They were not suffering from any chronic illness. After arrival, the subjects took 15e30 min rest in sitting posture prior to start the experiment with a normal room temperature of around 26
C. They were not prior exposed to heavy physical work. Then they were asked to put-on the experimental clothing: a half sleeve cotton shirt with their inner garments and a white cotton cloth, called dhoti, was double folded and plain wrapped around the waist to record the skin temperature easily with no active air movement in microenvironment area. They were bare feet during the experiment. The clothing insulation value (Icl) was estimated as 0.47 clo, where 0.04 clo: men’s briefs, 0.19 clo: half sleeved shirt, 0.24 clo: double folded dhoti. 2.2. Measurement of ambient parameters Experiment was conducted in the month of February and March at Bangalore. During that period, the average outdoor climate temperature was (26.4  3.9)
C (ranging from average daily minimum of 19.9
C and maximum of 32.8
C) and the average relative humidity was 46.95  18.8% and inside our campus the outdoor temperature was (18e32.0)
C with average of 25
C. Bangalore is having moderate climatic conditions throughout the year with average ambient temperature 33.5
C; solar radiation, 507 W m2; wind velocity, 4.9 m s1 [16]. In present study, indoor ambient temperature was varying from 21 to 33
C with an interval of 1
C except between 22
C and 21
C, it was 0.5
C. The whole experiment was conducted in two phases with a gap of at-least half a day to normalize their thermal response. First phase was conducted inside a conference room (Room A), where the ambient temperature was varied from 27
C to 21
C with gradual lowering the room temperature. The room ambient temperature was controlled by an air conditioner (LP-K3685QC e floor stand type AC, Capacity: 3.0 Ton, manufactured by LG Electronics) with an outlet (dimension of 0.55 m  0.32 m  1.8 m) situated at 1.6 m height, placed inside a conference room (2.5 m  5 m  3 m). The airflow rate of the cooler was 1054 m3/h. The minimum set temperature of the cooler was 16
C. This room was having insulated walls, false ceiling and PVC flooring. As, it was very difficult to increase the temperature beyond 27
C using air conditioner, therefore, the study of higher ambient temperature was conducted inside another laboratory room having almost same size (Room B), where the room temperature was adjusted between (28e33)
C in ascending fashion using one heating oven (model UNP 700, memmert) with heating area of 1040 mm  800 mm  500 mm, and two coil type cooking heaters (1000 W) and one pedestal fan not facing directly towards the subject (Fig.1). When the room ambient temperature reached to the set level, the oven and heaters were switched off and covered with a cloth to avoid the radiation from them and the fan was switched off temporarily and then the measurements were taken. Instead of using psychometric chamber, present study was conducted in a familiar natural environment as a general workplace for the subjects using auxiliary devices, where they could feel free and participate without being inside the enclosed chamber. Changing the ambient temperature from one level to the next usually took around 5 min, but 10 min was allotted and during this time subject relaxed in sitting. Then,

308

R. Maiti / Building and Environment 70 (2013) 306e317

Air conditioner

Heating Oven

D

B 1m

1m

Chair

1m

1.5 m Heater

E A

C

Fig. 1. A schematic layout of experimental room. In Room B, all the items were present except stand type air-conditioner. The length of the wall (AB) was 3 m and the heating oven was placed very near to BD wall. Heating oven (E) was placed 1.5 m from BD wall and the other one was placed at the same horizontal level and 1.0 m away from the subject. The fan was placed at 1 m distance from AB and AC walls. In Room A, which was a conference room where only air conditioner was placed near BD wall. The subject was standing 1.5 m in front and 1 m sideway from the air-conditioner.

ambient parameters were measured, which generally took around 5 min. Both Ambient/Air temperature (ta) and percentage of Relative Humidity (RH) were measured using Fluke Humidity meter (Fluke 971, Fluke Corporation, USA) with a RH range of 5e95% and accuracy of 2.5%, the measuring temperature range was 20 to 60
C (resolution of 0.1
C) with minimum accuracy of 0.5
C. Velocicalc air velocity meter with hot wire anemometery (Model 9515, TSI corporation, USA) was used to measure airflow (m/s) inside the room. This model was having measuring velocity range of 0e20 m s1 with resolution of 0.01 m s1, and air temperature range of 18e93
C with resolution of 0.1
C, having a telescoping measuring probe of 1 m length to reach to different heights. For PMV calculation, the ambient parameters at the position of the subject and at 1.2 m height from the ground were used. To study the variations at different vertical height levels at the same position, except globe temperature, other ambient parameters were measured with 0.5 m intervals from the floor up to the ceiling. The maximum difference between floor and the ceiling was observed as 0.8
C especially at higher ambient temperature. Air velocity inside the experimental rooms was obtained as (0.11  0.024) m s1. Relative humidity was almost constant everywhere and was recorded as (41.71  7.74)%. A black-globe thermometer was used to calculate mean radiant temperature (tr) using Equation (1). This thermometer consisted of a black sphere shell of 6.0 inch diameter and one thermometer was placed at the center as suggested by ISO standard 7726.

tr ¼




4 1:1  108 v0:6  a tg þ 273:15 þ tg  ta 3 D0:4

  1=4

 273:15

(1)

where, tg ¼ globe temperature (
C), va ¼ air velocity (m s1), ta ¼ air temperature (
C), D ¼ globe diameter (mm) and 3 ¼ emissivity (0.95 for a black globe).

subjective Thermal Sensation Vote (TSV) were recorded at each ambient temperature (ta). Skin temperature were recorded using RTD sensor (PT100) based temperature indicator (DC1000 Digital Temperature/Process Controllers, Honeywell International Inc.; with a range of (25 to 60)
C with resolution of 0.1
C and accuracy of 0.5
C) from eighteen different positions (Fig. 2) of a subject. The time required to get stable reading at each location was less than 5 s. From these segmental skin temperatures, mean skin temperature was calculated by following equation (Equation (2)), as suggested by Hardy and DuBois (1937) [23], a weighted method using constant weighting factors according to relative regional surface area of specific measuring sites.

Mean Skin Temperature ¼ 0:07  THead þ 0:35  TTrunk þ 0:14  TArm þ 0:05  THand þ 0:19  TThigh þ 0:13  TLeg þ 0:07  TFoot (2) where,

Head TemperatureðTHead Þ ¼ ðT7 þ T8Þ=2; Trunk TemperatureðTTrunk Þ ¼ ðT15 þ T16 þ T17 þ T18Þ=4; Arm TemperatureðTArm Þ ¼ ðT11 þ T12 þ T13 þ T14Þ=4; Hand TemperatureðTHand Þ ¼ ðT9 þ T10Þ=2;   Thigh Temperature TThigh ¼ ðT5 þ T6Þ=2;   Leg Temperature TLeg ¼ ðT3 þ T4Þ=2; Foot TemepratureðTFoot Þ ¼ ðT1 þ T2Þ=2; Chest TemperatureðTChest Þ ¼ ðT15 þ T17Þ=2; Back TemepratureðTBack Þ ¼ ðT16 þ T18Þ=2:

2.3. Measurement of skin temperature After recording the environmental parameters another 10 min was allotted for subjective adaptation to the room temperature at standing posture before recording body temperatures. During the experiment, regional skin temperature, oral temperature and

2.4. Measurement of core temperature and body temperature In present study, oral temperature was regarded as an index of core temperature. After recording regional skin temperatures, oral

R. Maiti / Building and Environment 70 (2013) 306e317

309

changed away from zero in either the positive or negative direction, PPD increased, indicated lesser percentage of population was thermally satisfied.

i h  PPD ¼ 100  95exp  0:03353PMV 4 þ 0:2179PMV 2

(5)

2.6. Recording of subjective Thermal Sensation Vote (TSV) After recording the oral temperature, subjective Thermal Sensation Vote was collected based on ASHRAE seven-point thermal sensation scale [1,2]. They are, respectively, 3 cold, 2 cool, 1 slightly cool, 0 just right (neutral), 1 slightly warm, 2 warm and 3 hot. Mean TSV values and the corresponding PPD values (replacing PMV with TSV in Equation (5)) were then fitted with Gaussian fit curve to obtain TSV-PPD curve. 2.7. Statistical analysis

Fig. 2. Identification of body position used for measuring regional skin temperatures.

temperature was measured by keeping a glass thermometer in a sublingual pocket under the tongue for about two minutes inside closed mouth. From the mean skin temperature and oral temperature, Body temperature (Tb) was calculated using following equation [24,25].

Tb ¼ 0:7Tcore þ 0:3Tsk

(3)

The effect of ambient temperature (ta) on the change of regional skin temperatures, mean skin temperature and core temperature values were analyzed by one-way ANOVA with post hoc testing by the Bonferroni method (alpha ¼ 0.001). Prior to that their normality was tested using ShapiroeWilk’s W test. Nonparametric tests (ManneWhitney test and KolmogoroveSmirnov test) were applied to evaluate the difference between calculated PMV and TSV responses at different ta with significance level of 0.05. Spearman’s rank correlation (two-tailed) test was used to explore the relation of TSV responses with regional skin temperature, mean skin temperature and core temperature. The relationship between dependent and independent variables was tested by linear and non-linear regression analysis, whereas stepwise regression analysis helped to automatically select the important independent variables. Multiple regression analysis was used to investigate the relationship between TSV response and different regional body temperatures.

2.5. Calculation of PMV and PPD

3. Results and discussion

PMV index was calculated based on Fanger’s steady state model [2,26]. Fanger separately introduced following relationship (Equation (4)) between skin temperature and the metabolic heat production (M) pointed below [26,27], used in tcl calculation. To study the thermal adaptive response, in present study, measured Tsk value was used in tcl calculation and results were compared.

3.1. Variation of skin temperatures and core temperature at different ta

Tsk ¼ 35:7  0:0275  M

(4)

After calculating PMV, PPD was estimated using following equation (Equation (5)), suggested by ASHRAE, 2005 [2]. As PMV

Both skin and core temperatures are effective indicators to evaluate subjective thermal sensation. Variation of regional skin temperature at different ta is given in Table 1, which shows a gradual increase with ta but at higher ta it slightly decreases mainly due to sweat evaporation, as visible sweat was observed in most subjects at ta ¼ 30
C. This result shows that Ttrunk (indicated by both Tchest and Tback) starts decreasing at ta ¼ 30
C. Tleg starts

Table 1 Changes (mean  sd) of regional skin temperature (
C) at different ta. ta (
C)

Thead (
C)

Tchest (
C)

Tback (
C)

Tuparm (
C)

Tlowarm (
C)

Thand (
C)

Tthigh (
C)

Tleg (
C)

Tfoot (
C)

21 21.5 22 23 24 25 26 27 28 29 30 31 32 33

33.16 33.23 33.51 33.87 34.03 34.02 34.28 34.4 34.71 34.94 35.03 35.17 35.17 34.92

33.56 33.79 33.88 34.01 34.02 34.16 34.23 34.28 34.78 34.89 34.88 34.68 34.36 34.13

33.49 33.64 33.84 33.9 33.91 34.08 34.22 34.32 34.59 34.75 34.67 34.42 34.17 33.78

32.12 32.23 32.56 32.75 32.98 33.12 33.19 33.36 33.31 33.56 33.75 33.99 33.97 33.84

32.3 32.55 32.81 33.03 33.17 33.29 33.44 33.56 33.71 34.03 34.22 34.3 34.35 34.23

32.71 32.84 33.26 33.44 33.47 33.73 33.74 33.79 33.86 34.15 34.31 34.44 34.44 34.13

32.09 32.24 32.44 32.77 32.77 33.06 32.96 33.23 33.19 33.56 33.62 33.67 33.48 33.30

32.24 32.36 32.62 32.99 32.93 33.23 33.29 33.69 33.55 33.85 33.91 33.87 33.71 33.58

31.85 32.05 32.42 32.74 32.99 33.01 33.34 33.24 33.34 33.84 33.78 33.79 33.79 33.65

(1.03) (1.0) (0.86) (0.8) (0.71) (0.77) (0.64) (0.77) (0.59) (0.61) (0.48) (0.53) (0.62) (0.96)

(0.65) (0.6) (0.58) (0.56) (0.63) (0.6) (0.64) (0.67) (0.52) (0.63) (0.69) (1.34) (1.39) (1.49)

(0.61) (0.53) (0.55) (0.68) (0.69) (0.74) (0.63) (0.64) (0.6) (0.74) (0.96) (1.26) (1.2) (1.31)

(0.84) (0.81) (0.71) (0.62) (0.64) (0.63) (0.72) (0.78) (0.81) (0.68) (0.84) (0.9) (0.91) (1.01)

(0.7) (0.78) (0.65) (0.57) (0.6) (0.73) (0.6) (0.65) (0.74) (0.64) (0.69) (0.85) (0.81) (0.92)

(0.84) (0.8) (0.72) (0.57) (0.64) (0.68) (0.55) (0.65) (0.77) (0.86) (0.8) (0.84) (0.74) (0.97)

(0.84) (0.92) (0.76) (0.81) (0.7) (0.81) (0.7) (0.67) (0.62) (0.50) (0.49) (0.58) (0.66) (0.93)

(0.64) (0.69) (0.71) (0.68) (0.67) (0.71) (0.62) (0.69) (0.62) (0.54) (0.52) (0.71) (0.82) (1.0)

(0.9) (0.89) (0.87) (0.81) (0.63) (0.79) (0.58) (0.74) (0.76) (0.78) (0.81) (0.8) (0.76) (0.96)

310

R. Maiti / Building and Environment 70 (2013) 306e317

lowering at ta ¼ 31
C, and Tlarm start lowering only at ta ¼ 33
C. Tuarm and Tthigh attain highest value at ta ¼ 31
C, whereas Thand and Tfoot remain almost at the same level at ta ¼ 31
C & 32
C and lowers at 33
C, and similarly in Thead. The difference between the mean core temperature and mean segmental skin temperature at different ta is given in Table 2. The result shows that head temperature is changing relatively more closely (2.109  0.419) with the core temperature than other body parts. Variations of both the core temperature, body temperature and mean skin temperature (Tsk) at different ta are given in Fig. 3. For ta from 24 to 26
C, the core temperature is changing slowly with 36.24  0.33
C (across all the subjects) and beyond this range of ta, the core temperature changes proportionally with ta. Similar to Tc, Tb follows similar response except in higher ta. Tsk gradually decreases in lowered ta region (21e23
C); whereas in higher ta, Tsk increases with ta and attains maximum value of 34.22
C at ta ¼ 30
C, then gradually decreases at higher ambient temperature region (31e33
C) as visible sweating observed at ta ¼ 30
C, but without excess dripping sweat even at ta ¼ 33
C. Analysis of the variance (ANOVA) results show that the regional skin temperature, mean skin temperature and core temperature at different ta are significantly different (alpha-0.001). Two different linear fit equations (Equations (6) and (7)) for lower (21e23)
C and higher (26e29)
C ta region on average Tsk response are plotted in Fig. 3, where lower slope in higher ta region and higher slope in lower ta region are observed.

For higher ta zone; Tsk ¼ 28:64631 þ 0:19019  ta   Adj: R2 ¼ 0:97867; p < 0:01

(6)

For higher ta zone; Tsk ¼ 26:21311 þ 0:3107  ta   Adj: R2 ¼ 0:96688; p < 0:01

(7)

Table 1 shows the temperatures of limb extremities have started lowering only at ta ¼ 33
C, which indicates that limb extremities are least warm-sensitive segment [28]. Patterson et al. have shown that after humid heat acclimatization on a non-acclimatized person causes greater relative increase in rate of sweating (mrsw) in the trunk region compared to other body regions except in forearm, which does not support the hypothesis of a trunk-to-limb sweat redistribution during acclimatization [29]. They also have observed that chest is having lowest sweating threshold temperature than any other body regions, which causes earlier onset of sweating in the trunk region, as observed in the present study. Nadel et al. have reported that thermal sensitivity with mean area weighted skin temperature, determining the regulation of sweating rate at different body regions are as follows: face, chest and back, abdomen, upper legs, upper arm, lower leg and lower arm [30], which is similarly reported by Cotter and Taylor (2005) [28]. Again, Lee et al. have reported that in case of tropical natives regional warm sensitivity descends in following order like head, trunk, arm/ hand, and calf/foot [31]. Similar response is observed in present study, except the leg skin temperature, which started falling at ta ¼ 30
C, as these subjects have worn dhoti, a simple drapery, during this experiment, which probably caused lower external conductance of the body heat and earlier onset of leg sweating. This

Fig. 3. Variations (Mean  sd) of Tsk, core (oral) temperature and body temperature at different ta. Active sweating observed at ta ¼ 30
C. Reference Tsk (using Equation (10)) is indicated to highlight the actual variation of Tsk at different ta.

difference in warm sensitivities at different body regions occurs only with lower warm stimulus, but as the intensity of the stimulus increases more uniform sweating response occurs in various body parts [32]. In present study, the head skin temperature is found to be relatively more thermal sensitivity than other body parts and changing more closely with core temperature (Table 2); hence the heat dissipation through head is significantly higher with the normalized body surface area than any other body region [28,33]. Machado-Moreira et al. [34] have observed lower sweat rate from the top of the head in hairy subject and in present study due to hair on the top of the head moisture started accumulating on the head skin and leading to possible hydro-meiotic suppression of sweat evaporation and caused less evaporative heat loss from head at higher ta region. Gagge and Gonzalez have reported that sweating occurs after the core temperature increases of about 0.2e0.3
C from a base line core temperature of 37.0  0.5
C [35]. Fig. 3 shows the core temperature changes from constant level (36.24  0.33
C) to 36.72  0.34
C (Tcore at ta ¼ 30
C), then Tsk starts decreasing due to sweat. Cheuvront et al. have estimated this thermoregulatory sweating threshold temperature at which sweat rate (mrsw) exceeds baseline (threshold) of 0.06 mg cm2 min1 and measured as 36.97  0.3
C using Segmented Regression method [36], whereas others have reported that below a critical core temperature (i.e. sweating threshold temperature), no active sweating occur regardless of the level of skin temperature even up to 40
C [37,38]. Wyndham and Atkin [37] have obtained this sweating threshold temperature of 36.5
C from acclimatized subjects. Again Lopez et al. have recorded the sweating and shivering threshold in man as 37
C and 35.6
C respectively [39]. In present study, Tcore at ta ¼ 30
C is recorded as 36.72  0.34
C when visible sweat occurs, which is even slightly lower than reported by

Table 2 Overall difference (mean  sd) of core (oral) temperature with regional skin temperature.

Mean SD

Tc  Thead

Tc  Tchest

Tc  Tback

Tc  Tuparm

Tc  Tlowarm

Tc  Thand

Tc  Tthigh

Tc  Tleg

Tc  Tfoot

2.109 0.419

2.167 0.353

2.3 0.4

3.226 0.317

2.926 0.375

2.69 0.292

3.398 0.301

3.154 0.367

3.297 0.412

311

2.375 (0.54) 2.88 (0.26) 34.04 (0.73) 1.975 (0.42) 2.28 (0.29) 34.18 (0.72) 1.55 (0.6) 1.84 (0.16) 34.23 (0.49) 1.225 (0.48) 1.48 (0.14) 34.19 (0.72) 0.9 (0.38) 1.09 (0.15) 33.93 (0.44) 0.775 (0.42) 0.84 (0.13) 33.79 (0.39) 0.25 (0.44) 0.45 (0.15) 33.6 (0.44) 0.025 (0.16) 0.167 (0.16) 33.56 (0.46) 0.025 (0.28) 0.06 (0.17) 33.34 (0.51) 0.375 (0.54) 0.29 (0.15) 33.33 (0.42) 1.1 (0.5) 0.59 (0.17) 32.88 (0.49) TSV PMV (Tsk) Tsk

1.525 (0.55) 0.62 (0.16) 32.72 (0.46)

0.75 (0.49) 0.52 (0.14) 33.11 (0.41)

32
C 31
C 30
C 29
C 28
C 27
C 26
C 25
C 24
C 23
C 22
C 21.5
C

where, heat of vaporization from sweat at 35
C is 0.68 W h gm1, Tbset and Tskset are the set points of body temperature and skin temperature of 36.49
C and 33.7
C respectively and this mrsw depends on skin blood (Skbf) and is proportional to deviations in core and skin temperatures set points as Skbf allows passive heat conduction from the core compartment to the skin. The above equation is based on the assumption that the rate of sweating (mrsw) is increased by an average of 170 g/h m2 with increase of Tb by 1.0
C. In present study, mean values of Tsk e 33.6  0.44
C and 33.79  0.39
C occur at ta ¼ 26
C and 27
C respectively (Table 3), therefore, Tskset of 33.7
C may be not sufficient to consider in calculating mrsw. As well, in present study, the mean body temperature at 33
C is obtained as (33.11  0.32)
C. So, consideration of Tbset of 36.49
C is sufficiently higher for these data. Lee et al. have shown the sweat output per gland is 35.4% higher in temperate subjects (11.01  1.47 mg min1 single gland1) compared to tropical native subjects (8.13  1.26 mg min1 single gland1) [44]. Therefore, the factor 170 may need to be modified for Indian context, as due to tropical adaptation sweat rate decreases. Recent literature is limited to explain the actual sweating response from Indians. Thus required further study on sweat adaptation for Indians and reestablish the relations between metabolic rate, sweating rate and skin temperature with the thermal sensation response. Wyndham and Atkin have shown in acclimatized subjects sweat rate does not increase markedly until the skin temperature reaches beyond 33
C [37]. DiPasquale et al. [46] have revealed that regional skin temperature above approximately 32
C predominantly affects local active sweat secretion due to neurotransmitter Acetylcholine (Ach) release; and local skin temperature below this level affects local sweat production by altering glandular sensitivity controlled by core temperature. In present study, maximum regional skin temperature at different parts of the body before sweating is obtained >32
C (Table 1), even measured regional skin temperature at different ta (21e 33
C) almost remained higher than 32
C. This higher cutaneous thermal threshold for sweating is also reported by Lee et al., where they have mentioned tropical natives show on average of 3.3
C, 3.5
C, 4.2
C and 7.3
C higher values for warm, hot, cool, and cold sensations respectively compared to temperates and with local skin heating, sweat has not started even with a rise in local skin temperature up to 43.8
C [47]. In tropical natives, this higher threshold temperature causes more dry heat loss (radiation and

21
C

(8)

Table 3 Comparative results of thermal sensation response obtained from PMV calculation and TSV recorded response at different ta along with the change in Tsk.



   T T mrsw g=hm2 ¼0:68 170$ðTb Tbset Þ$exp sk skset 10:7

33
C

Cheuvront et al. [36] and Mekjavi c and Bligh [40]. This result indicates a lowered sweat threshold temperature in these tropical natives as a result of adaptation. In present study, with continued positive heat load, in these subjects sweat production increases in proportion to the increase in core temperature in the form of more and more profuse sweat, without dripping of sweat as a response of tropical adaptation. This is corroborated with previous results, where several authors have reported that prolonged tropical natives show heat tolerance adaptation with suppressed sweating and with larger wetted area without dripping sweat promotes improved evaporation [41e44]. Unfortunately, this dripping sweat makes no contribution to cool the body. Lee et al. have proposed this suppressed sweating is due to the blunted sudomotor sensitivity of the sweat gland to Acetylcholine (Ach) [44]. Nevertheless, this sweat rate for evaporative heat loss (mrsw) can be calculated from the deviation of set points of mean body temperature (36.49
C) and skin temperature (33.7
C) (Equation (8)), as given below [45].

2.675 (0.47) 3.47 (0.29) 33.82 (0.86)

R. Maiti / Building and Environment 70 (2013) 306e317

312

R. Maiti / Building and Environment 70 (2013) 306e317

convection) in tropical natives compared to temperate population [48] as the amount of heat dissipation is proportional to the temperature gradient of (Tsk  ta) and requires less sweat production and evaporation for same temperature gradient due to tropical adaptation. Many authors have studied on short-term hot-humid tropical on non-acclimatized person and concluded that acclimatization causes overall increase in sweat rate secretion (mrsw) and alters distribution of sweating area but not a generalized redistribution toward the limbs [29]. This increased secretary capacity of sweat glands due to stronger sweating drive but not resulting from a decrease in the central temperature reference [49,50]. These above adaptive response results should not mislead as very longterm (2e13 years) tropical adaptation response [43] and on the contrary different authors have proved that tropical inhabitants are having lowered core and skin temperature [44,51] due to a lower metabolic rate resulting from lower thyroxin secretion [51]. In spite of that, this level of tropical adaptation response can have slight variation due to short term seasonal acclimatization [52]. Interestingly, Fig. 3 shows that mean skin temperature (Tsk) response shows different slope at lower ta (21e23)
C and higher (26e29)
C ta regions of 0.3107 and 0.19 respectively to compensate temperature gradient between skin and air temperature. Davis et al. have reported that the skin vasoconstriction begins when the skin temperature falls below 33
C (95
F) [53]; but here it is observed Thead, Tchest, Tback were having >33
C even at ta ¼ 21
C. This higher temperature gradient of Tsk  ta would cause increased dry heat loss at lower temperature in these subjects. Nevertheless, Lee et al. [47] have reported heat acclimatization in tropical native shows a blunted cold sensitivity response, but they are unable to explain the methodology. From these results, it is proposed that tropical population is more heat tolerant in higher ambient temperature and in lower ambient temperature, these people shows an enhanced cold-induced vasoconstriction instead of blunted coldsensitivity response.

3.2. Relationship between TSV and skin temperature and core temperature Significant positive correlations exist between TSV and mean skin temperature (Spearman rho ¼ 0.58158, p < 0.001) and core temperature (Spearman rho ¼ 0.671848, p < 0.001) and body temperature (Spearman rho ¼ 0.76849, p < 0.001). Multiple regression analysis (Equation (9)) shows following relationship between TSV and Tsk, Tc and Tb, whereas stepwise regression analysis shows only Tb is important parameter in determining TSV (Equation (10)). Therefore, Tb is better predictor for TSV than Tsk and Tc, which is similarly reported by Nielsen and Nielsen [54].

TSV ¼ 78:6583  1:0923  Tsk  2:4423  Tc þ 5:7586 Tb   R2 ¼ 0:5771; Adj: R2 ¼ 0:5748 TSV ¼ 78:3711 þ 2:2184  Tb



R2

 ¼ 0:5765 2

(9)

Table 4 ANOVA analysis results on the effect of different regional skin temperature on TSV. Model

DF

R2

Intercept 1 Model 9 0.493 Head T 1 0.051 Foot T 1 0.012 Leg T 1 0.003 Thigh T 1 0.000 Hand T 1 0.004 LArm T 1 0.031 UArm T 1 0.000 Back T 1 0.003 Chest T 1 0.005 Error 550 0.507 Total (adj) 559 1.000

SS

MS

181.716 487.082 50.773 11.705 3.388 2.055E-02 3.516 30.572 9.265E-03 2.611 4.435 500.202 987.284

181.716 54.12 50.773 11.705 3.388 2.055E-02 3.516 30.572 9.265E-03 2.611 4.435 0.909 1.766

F-ratio p-value % contribution 59.508 55.827 12.87 3.725 0.023 3.866 33.615 0.010 2.871 4.876

0.0000 0.0000 0.0004 0.0541 0.8806 0.0498 0.0000 0.9196 0.0907 0.0276

5.14 1.18 0.34 2.08E-05 0.35 3.1 9.38E-06 0.26 0.45

TSV ¼ 30:2808 þ 0:4296  THead þ 0:2145  TFoot þ 0:1612  TLeg þ 0:0119  TThigh  0:1646  THand þ 0:6028  TLArm  0:0083  TUArm  0:1396  TBack  0:1853  TChest (11) Linear stepwise regression analysis is used to optimize the linear combination of important parameters in the model with ensured parsimony. Table 1 shows a gradual increase of regional skin temperature with ta except at higher ta region, where it decreases mainly due to sweat evaporation. However, all different regional skin temperatures don’t start decreasing at a particular ta. Therefore, when the stepwise regression is done for ta (21e29)
C, then TFoot, TLeg, TBack, THead (R2 ¼ 0.421081, Sqrt(MSE) ¼ 0.7289542) are identified as significant parameters in determining TSV. For ta (30e33)
C only TThigh and TLArm (R2 ¼ 0.119771, Sqrt(MSE) ¼ 0.6250933) are shown as significant parameters, as all other parameters changes both in positive and negative direction. This kind of variation occurs due to the fact that regional skin temperature response is modified by peripheral blood flow, uneven clothing distribution, sweating etc., and changes both in positive and negative direction systematically based on ta. Linear stepwise regression analysis for entire ta (i.e. 21e33
C) shows that only THead, TFoot, TLArm and TChest are significant contributors in TSV calculation (p < 0.0001) (Table 5). While calculating correlation between TSV and different regional body temperature, it is observed that the Spearman rho for TFoot e 0.56785; TLeg e 0.53112; TThigh e 0.49141; THand e 0.55357; TLArm e 0.63653; TUArm e 0.551; TBack e 0.32065; TChest e 0.40228; THead e 0.62657. This result shows that TBack and TChest are having lower correlation with TSV as TTrunk is more sensitive to sweat response in higher ta, as well trunk region is having higher local thermal adaptation capacity. THead shows higher correlation as the head region is having lower local thermal adaptation than other body region [56] as head is a very important part for heat dissipation and THead is varying relatively more closely with core temperature in comparison with other body regions (Table 2). 3.3. Thermal Sensation Vote (TSV) and PMV values at different ta

(10)

Multiple regression analysis results (R ¼ 0.4934, Adj. R2 ¼ 0.4851) shows the following equation between TSV response and different regional body temperatures (Equation (11)). Analysis of Variance result shows that THead contributes highest value of 5.14% of total variance, whereas TFoot and TLArm contributes 3.1% and 1.18% respectively (Table 4), which shows head is more thermal sensitive [55].

A comparative result of subjective Thermal Sensation Vote (TSV), calculated PMV values and mean skin temperature (Tsk) at different ta are given in Table 3, where ta ¼ 28
C and ta ¼ 31
C are sensed as slight warm (0.925  0.35) and warm (1.975  0.423) respectively. Similarly, the corresponding ta for slight cool sensation is calculated as (22 þ 21.5)/2
C ¼ 21.75
C and the corresponding average TSV value is 0.95. From slightly warm to warm sensation, the subject starts sweating and Tsk decreases.

R. Maiti / Building and Environment 70 (2013) 306e317 Table 5 Summary of Stepwise Linear Multiple Regression analysis results on the effect of different regional skin temperature on TSV response. The models are also presented herewith. Model

Parameters

Std. coeff.

R2

Adj R2

Sqrt(MSE)

a b

c6 c6 c10 c6 c9 c10 c2 c6 c9 c10

0.621 0.429 0.294 0.473 0.192 0.373 0.196 0.385 0.204 0.325

0.385 0.435

0.385 0.433

1.043 1.001

0.459

0.457

0.98

c

d

0.481

0.477

0.961

Model: a: TSV ¼ 27.9048 þ 0.850*TLArm. b: TSV ¼ 32.2873 þ 0.3839*THead þ 0.587540270986097*TLArm. c: TSV ¼ 28.5015 þ 0.4860*THead þ 0.6474*TLArm  0.2713*TChest. d: TSV ¼ 30.2635 þ 0.5275*TLArm  0.2875*TChest þ 0.4239*THead þ 0.2555*TFoot.

Thermal Sensation Vote (TSV) is found to have a good linear relationship with ambient temperature (ta) (Equation (12)). Although, linear regression shows a very good fit for the entire ta range, but shows larger discrepancy in lower ta region (red-dashed line in Fig. 4). Linear regression fitting separately in higher and lower ta regions (Equations (13) and (14)) provides better fit, which are presented by solid lines in Fig. 4. However, non-linear fitting of TSV for entire ta region can be slightly improved by non-linear curve fitting (Equation (15)). However, for PMV results, 2nd degree polynomial fitting (Fig. 5) shows much better response than linear fitting (Equations (16) and (17)). Neutral temperature is calculated as the corresponding ta where thermal sensation value is zero. The neutral temperatures from TSV response are obtained as 24.83
C and 24.23
C from linear and non-linear fittings respectively (Equations (12) and (15)), whereas from PMV response these are obtained as 23.91
C and 24.41
C from linear and non-linear fittings respectively (Equations (16) and (17)).

Fig. 4. Linear relationship between Thermal Sensation Vote (TSV) responses at different ta. Linear relationship of TSV for entire ta is represented by red dashed line. Black solid lines denote the linear relationships of TSV in higher and lower ta regions separately. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

313

TSV ðfor entire ta rangeÞ ¼ 7:93888 þ 0:31978  ta   Adj: R2 ¼ 0:98576; p < 0:001

(12)

TSVðLower ta region 21  23 o CÞ ¼ 13:25 þ 0:56286  ta   Adj: R2 ¼ 0:93755 (13) TSV ðHigher ta region 26  33 o CÞ ¼ 8:37308 þ 0:33381  ta   Adj: R2 ¼ 0:98379 (14) TSV ðfor entire ta rangeÞ ¼ 388:04884 þ 56:28795 ta  3:06041  ta2 þ 0:07369 

ta3  6:59524E  4  ta4  Adj: R2 ¼ 0:99526; p < 0:001

(15)

PMV ¼ 7:85062 þ 0:32835  ta   Adj: R2 ¼ 0:96695; p < 0:001

(16)

PMV ¼ 4:46749  0:61048  ta þ 0:01751  ta2   Adj: R2 ¼ 0:99831; p < 0:001

(17)

ManneWhitney test shows that except ta ¼ 23
C and 30
C, all other ta, calculated PMV and TSV responses are significantly different (alpha-0.001), whereas KolmogoroveSmirnov test shows they are significantly different at all ta (alpha-0.001), as unlike ManneWhitney test (test the difference in means, median, average rank), the KolmogoroveSmirnov test is also sensitive to the difference in the general shapes of the distribution in two samples (i.e. difference in dispersion, skewness etc.). Calculated PMV values and TSV responses from all the subjects at different ta are presented by a scattered plot (Fig. 6). A linear regression analysis shows a very good relation (R2 ¼ 0.83466) between PMV and TSV but the linear fit line of PMV-TSV response at different ta (solid line in Fig. 6) is having different slope than the centerline and remains always

Fig. 5. Both linear (black solid line) and non-linear fit (black dotted line) of calculated PMV values at different ta.

314

R. Maiti / Building and Environment 70 (2013) 306e317

PMV model as well from TSV-PPD curve. In present study, it is observed that mean PMV values exceeds beyond þ3 limit, whereas maximum mean TSV is obtained as 2.675 (Table 3). Doherty and Arens have reported that the discrepancies between predicted and actual thermal sensation can be as large as 1.3 scale units [45]. According to Humphreys and Nicol, the ASHRAE scale should not extend beyond 3 scale unit and in population where a grandmean comfort vote approaches either extreme of the scale or exceeds then it is unlikely to have captured the true response of the people by PMV results [58]. A good Gaussian fit curve (magenta line) is obtained in TSV-PPD curve (Fig. 8) and the corresponding equation is given below (Equation (18)). From both TSV-PPD and PMV-PPD response (redline) and the linear relationship between thermal sensation responses (PMV and TSV) at different ta, thermal comfort zone are estimated as given in Fig. 8.

PPD ¼ 110:53  103:148*e0:24ðTSVþ0:0077Þ   Adj R2 ¼ 0:99658; p < 0:001 Fig. 6. Scattered plot of TSV and PMV values. Centerline is represented by black dotted line and the linear fit of the data is denoted by black solid line.

below the centerline; although, skin temperature is used in PMV calculation to include subjective adaptive response. This shows that PMV overestimates the actual thermal sensation response [6,57], which may be due to the fact that the heat balanced PMV equation does not account for people’s adaptation response. Again the difference between PMV and TSV (PMV-TSV) is plotted at different ta, (Fig. 7) to test whether this difference is having any specific trend at different ta. It is observed that both the linear and non-linear regression analysis shows very poor relationship of this (PMVTSV) with ta (Adj. R2 ¼ 0.0019 and 0.16396 respectively), which interprets that the difference between calculated PMV values and TSV responses are not biased with ta. 3.4. Estimation of thermal comfort temperature

2

(18)

It is observed that in TSV-PPD graph, 0.5  TSV  þ0.5 corresponds to 13.47% PPD level and 10% PPD level is corresponding to 0.35  TSV  þ0.35 and 0.5  PMV  þ0.5 (Fig. 8). With 0.5  TSV  þ0.5, the comfort temperature zone for these subjects is estimated as (23.25e26.32)
C with neutral temperature of 24.83
C, whereas for (0.5  PMV  þ0.5) it is corresponding to (22.46e25.41)
C with neutral temperature of 23.91
C. For (1  TSV  þ1), the comfort zone is obtained (21.73e27.92)
C. Therefore, estimated comfort zone using PMV model produce lower temperature range than that obtained from TSV response. Lee et al. have shown that the cutaneous interthreshold sensory zone, the zone between thermal thresholds for ‘slightly warm’ and ‘slightly cool’ sensations is wider in tropical natives than temperate group [47]. However, Fanger’s (1970) original studies were conducted using white, college-age participants [27]. The PMV model resulting from these studies might not, therefore, be equally valid for other populations, as is observed in present study. However, using adaptive comfort standard [59], the

From the calculated PMV values, PPD curve is generated to evaluate thermally comfort air temperature for Indians based on

Fig. 7. Difference between PMV and TSV (viz. PMV-TSV) at different ta along with linear and non-linear fit curves.

Fig. 8. Comparative thermal sensation results obtained using calculated PMV values and actual subjective thermal sensation response with corresponding ta. Linear fit curves of PMV (blue line) and TSV (red line) at different ta are considered for estimating thermal comfort zone. Calculated PMV-PPD% response is presented by black line. Gauss fit curve for TSV-PPD% is plotted in magenta, which is extended with dotted curve. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

R. Maiti / Building and Environment 70 (2013) 306e317

optimum temperature is obtained as 25.55
C (with mean outdoor temperature of 25
C) with 5.5
C for 90% acceptability, produce a comfort zone of (20e31)
C for this population. Based on a field study at Hyderabad, a metro city of India, the comfort band (voting within 1 and þ1), was obtained as 26e32.45
C, with the neutral temperature at 29.23
C in a naturally ventilated apartments, where the indoor temperature was varying between 26.6
C and 42
C and the participants were having average age of 40.12 years and 42 years for male and female respectively [19]. However, present study was conducted at Bangalore, which is having moderate temperature throughout the year and inside our campus the outside temperature was (18e32.0)
C with an average of 25
C. This kind of discrepancy between PMV and TSV of estimated comfort zone can be explained by the adaptive response during seasonal variation, age of the participants and behavioral adaptation. It is reported in the literature, that seasonal variation can cause a shifted thermal comfort zone with a shift in neutral temperature of w3
C [57,60]. Lee et al. have revealed tropical natives were less sensitive to detect warmth than temperate natives [31]. Again, the elderly people generally have decline in thermal sensitivity and increased body temperature threshold for the onset of sweating [61] and they manipulate much less precisely the ambient temperature [62], therefore their thermal preference is generally having a wider ambient temperature fluctuation with poor reproducibility of preferred ambient temperature [63]. These older group also require higher ta for comfort compared to younger counterpart [64] and require more intense thermal stimulus to elicit a thermal behavioral response at home environment [65]. However, in relation to thermal sensation, behavioral adaptation with respect to cultural adjustments, lifestyle etc. at different parts of India is reported by the researchers [66,67]. From this result, it is shown that PMV result is always higher than the actual feeling of the occupants on thermal environment and therefore, national standard needs to be modified towards better energy-efficient buildings design.

315

Fig. 10. Comparative results of actual PMV-PPD% response and TSV-PPDnew% response to highlight the positive skewness of the actual subjective response in higher ta regions, which is further extended with dotted curve.

data, given below (Equation (19)). Here it is observed that the center point of the curve is shifted towards positive to 0.53175.

PPD ¼ 112:35953  107:395*e0:173587*ðx0:53175Þ

2

(19)

This relation is further improved by asymmetric Gaussian fit (Adj. R2 ¼ 0.97496) and a new Predicted Percentage of Dissatisfaction (PPDnew) curve is generated (Fig. 10), where the center point is attained at 0.02514 and the width parameters of the Gaussian fit (viz. 2*sigma) are obtained 1.01057 and 2.01381 respectively. Fig. 10 shows a positive skewed curve, which indicate the longer tail in the higher ta region. This indicates that Indians are more heat tolerant in higher ta than lower ta region as repeated warm stimuli only increases the activity of warm receptors.

3.5. A new relation between percentage of dissatisfaction and TSV To study the thermal response for Indians, the vote obtained for different TSV response is plotted in Fig. 9. The relationship between percentage dissatisfied and the thermal response is obtained by a symmetric Gaussian fit curve (Adj. R2 ¼ 0.92111) on the normalized

Fig. 9. Histogram of number of responders at different TSV values.

4. Conclusion This study gives an overview of adaptive thermal response from Indian subjects based on laboratory experiment. Following interesting results are obtained from this study are given below.  Thermal adaptive changes are observed in distribution of regional skin temperature, higher cutaneous threshold temperature for sweating and lowered sweating threshold temperature.  It is also shown calculation of sweat rate for these population demands new research.  Interestingly, mean skin temperature (Tsk) response shows different slope at lower ta (21e23)
C and higher (26e29)
C ta regions of 0.3107 and 0.19 respectively, which can not be explained by blunted sensitivity of neurotransmitter in cold thermal conditions.  Although, significant positive correlations exist between TSV and Tsk, Tc and Tb, but Tb is better predictor of TSV response than Tsk and Tc.  Among different regional body temperature THead shows prominent role in determining TSV response.  Calculated PMV values and TSV responses are significantly different and PMV overestimates TSV values. Calculated PMV values goes beyond þ3 limit at higher ta.  From TSV-PPD graph, it is seen that 0.5  TSV  þ0.5 corresponds to 13.47% PPD level and 10% PPD level is corresponding

316

R. Maiti / Building and Environment 70 (2013) 306e317

to 0.35  TSV  þ0.35 and 0.5  PMV  þ0.5. With 0.5  TSV  þ0.5, the comfort temperature zone for these subjects is estimated as (23.25e26.32)
C with neutral temperature of 24.83
C, whereas for (0.5  PMV  þ0.5) it is corresponding to (22.46e25.41)
C with neutral temperature of 23.91
C. For (1  TSV  þ1), the comfort zone is obtained (21.73e27.92)
C.  A new TSV-PPD relation (PPDnew) shows an asymmetric distribution of hot-cold thermal sensation response in Indians, where subjects are more heat tolerant than cold thermal sensation.

Acknowledgment I would like to thank to all the people who helped in data collection in this study and volunteered in this study.

References [1] ASHRAE. ANSI/ASHRAE standard 55-1992. Thermal environmental conditions for human occupancy. Atlanta: American Society of Heating, Refrigerating, and Air-Conditioning Engineers Inc.; 1992. [2] ASHRAE. ASHRAE handbook of fundamentals. Atlanta: American Society of Heating Refrigeration and Air-Conditioning Engineers Inc.; 2005. [3] ISO 7730. Ergonomics of the thermal environment e analytical determination and interpretation of thermal comfort using the PMV and PPD indices and local thermal comfort criteria. Switzerland: International Organisation for Standardization; 2005. [4] Fanger PO. Thermal comfort, analysis and applications in environmental engineering. New York: McGraw Hill; 1972. [5] Bouden C, Ghrab N. An adaptive thermal comfort model for the Tunisian context: a field study results. Energy Build 2005;37(9):952e63. [6] Brager G, de Dear R. Thermal adaptation in the built environment: a literature overview. Energy Build 1998;27:83e90. [7] Wong LT, Mui KW. Efficiency assessment of indoor environmental policy for air-conditioned offices in Hong Kong. Appl Energy 2009;86:1933e8. [8] de Dear RJ, Brager GS. Thermal comfort in naturally ventilated buildings: revisions to ASHRAE standard 55. Energy Build 2002;34(6):549e61. [9] Wong NH, Feriadi H, Lim PY, Tham KW, Sekhar C, Cheong KW. Thermal comfort evaluation of naturally ventilated public housing in Singapore. Build Environ 2002;37:1267e77. [10] Nicol F. Adaptive thermal comfort standards in the hot-humid tropics. Energy Build 2004;36:628e37. [11] Wijewardane S, Jayasinghe MTR. Thermal comfort temperature range for factory workers in warm humid tropical climates. Renew Energy 2008;33: 2057e63. [12] Hwang RL, Cheng MJ, Lin TP, Ho MC. Thermal perceptions, general adaptation methods and occupant’s idea about the trade-off between thermal comfort and energy saving in hot-humid regions. Build Environ 2009;44:1128e34. [13] Yao R, Li B, Liu J. A theoretical adaptive model of thermal comfort e adaptive predicted mean vote (aPMV). Build Environ 2009;44:2089e96. [14] de Dear RJ, Brager GS. Developing an adaptive model of thermal comfort and preference. ASHRAE Trans 1998;104(1):1e18. [15] Hwang RL, Lin TP, Kuo NJ. Field experiments on the thermal comfort in campus classrooms in Taiwan. Energy Build 2006;38:53e62. [16] Singh MC, Garg SN, Jha R. Different glazing systems and their impact on human thermal comfortdIndian scenario. Build Environ 2008;43(10):1596e 602. [17] Sharma MR, Ali S. Tropical Summer Index e a study of thermal comfort of Indian subjects. Build Environ 1986;21(1):11e24. [18] Mallick FH. Thermal comfort and building design in the tropical climates. Energy Build 1996;23:161e7. [19] Indraganti M. Thermal comfort in naturally ventilated apartments in summer: findings from a field study in Hyderabad, India. Appl Energy 2010;87(3):866e 83. [20] BIS. National building code. Bureau of Indian Standards; 2005. [21] Bulaco CF, Frank SM, Raja SN, Tran KM, Goldstein DS. Relative contribution of core and skin temperatures to thermal comfort in humans. J Therm Biol 2000;25:147e50. [22] DuBois D, DuBois DF. A formula to estimate the approximate surface area if height and weight be known. Arch Intern Med 1916;17:863e71. [23] Hardy JD, DuBois EF. The technic of measuring radiation and convection. J Nutr 1937;15(5):461e75. [24] Hardy JD, DuBois EF. Basal metabolism, radiation, convection, and evaporation at temperature from 22
to 35
C. J Nutr 1938;15:477e92. [25] Stolwijk JAJ, Hardy JD. Partitional calorimetric studies of responses of man to thermal transients. J Appl Physiol 1966;21:967e77.

[26] ISO. ISO/TR 11079 Evaluation of cold environmentsddetermination of required clothing insulation (IREQ). Geneva, Switzerland: International Organization for Standardization; 1993. [27] Fanger PO. Thermal comfort. Copenhagen: Danish Technical Press; 1970. [28] Cotter JD, Taylor NA. The distribution of cutaneous sedomotor and alliesthesial thermosensitivity in mildly heat-stresses humans: an open-loop approach. J Physiol 2005;565(1):335e45. [29] Patterson MJ, Stocks JM, Taylor NA. Humid heat acclimation does not elicit a preferential sweat redistribution toward the limbs. Am J Physiol Regul Integr Comp Physiol 2004;286(3):R512e8. [30] Nadel ER, Mitchell JW, Stolwijk JA. Differential thermal sensitivity in the human skin. Pflügers Arch 1973;340:71e6. [31] Lee JY, Saat M, Chou C, Hashiguchi N, Wijayanto T, Wakabayashi H, et al. Cutaneous warm and cool sensation thresholds and the inter-threshold zone in Malaysian and Japanese males. J Therm Biol 2010;35(2):70e6. [32] Stevens JC, Marks LE, Simonson DC. Regional sensitivity and spatial summation in the warmth sense. Physiol Behav 1974;13:825e36. [33] Rasch W, Samsom P, Cote J, Cabanac M. Heat loss from the human head during exercise. J Appl Physiol 1991;71(2):590e5. [34] Machado-Moreira CA, Wilmink F, Meijer A, Mekjavic IB, Taylor NA. Local differences in sweat secretion from the head during rest and exercise in the heat. Eur J Appl Physiol 2008;104:257e64. [35] Gagge AP, Gonzalez RR. Mechanism of heat exchange: biophysics and physiology. In: Fregly MJ, Biatteis CM, editors. Handbook of physiology, volume 1, Sect 4: environmental physiology. New York: Oxford University Press; 1996. p. 45e84. [36] Cheuvront SN, Bearden SE, Kenefick RW, Ely BR, DeGroot DW, Sawka MN, et al. A simple and valid method to determine thermoregulatory sweating threshold and sensitivity. J Appl Physiol 2009;107:69e75. [37] Wyndham CH, Atk1ns AR. A physiological scheme and mathematical model of } gers Arch 1968;303:14e30. temperature regulation in man. Pflu [38] Nadel ER, Bullard RW, Stolwijk JAJ. Importance of skin temperature in the regulation of sweating. J Appl Physiol 1971;31:80e7. [39] Lopez M, Sessler DI, Walter K, Emerick T, Ayyalapu A. Reduced sweating threshold during exercise-induced hyperthermia. Pflügers Arch 1995;430(5): 606e11. [40] Mekjavi c IB, Bligh J. Core threshold temperatures for sweating. Can J Physiol Pharmacol 1989;67(9):1038e44. [41] Matsumoto T, Kosaka M, Yamauchi M, Tsuchiya K, Ohwatari N, Motomura M, et al. Study on mechanisms of heat acclimatization due to thermal sweatingdcomparison of heat tolerance between Japanese and Thai subjects. Trop Med 1993;35:23e34. [42] Lee JB, Matsumoto T, Othman T, Kosaka M. Suppression of the sweat gland sensitivity to acetylcholine applied iontophoretically in tropical Africans compared to temperate Japanese. Trop Med 1997;39:111e21. [43] Lee JB, Bae JS, Lee MY, Yang HM, Min YK, Song HY, et al. The change in peripheral sweating mechanisms of the tropical Malaysian who stays in Japan. J Therm Biol 2004;29(7e8):743e7. [44] Lee JB, Bae JS, Matsumoto T, Yang HM, Min YK. Tropical Malaysians and temperate Koreans exhibit significant differences in sweating sensitivity in response to iontophoretically administered acetylcholine. Int J Biometeorol 2009;53(2):149e57. [45] Doherty TJ, Arens E. Evaluation of the physiological bases of thermal comfort models. ASHRAE Trans 1988;94(1):1371e85. [46] DiPasquale DM, Buono MJ, Kolkhorst FW. Effect of skin temperature on the cholinergic sensitivity of the human eccrine sweat gland. Jpn J Physiol 2003;53(6):427e30. [47] Lee JY, Bakri I, Toramoto S, Tochihara Y. Cutaneous thermal thresholds of tropical indigenes residing in Japan. J Therm Biol 2011;36:461e8. [48] Hori S. Adaptation to heat. Jpn J Physiol 1995;45:921e46. [49] Fox RH, Goldsmith R, Kidd DJ, Lewis HE. Blood flow and other thermoregulatory changes with acclimatization to heat. J Physiol (Lond) 1963;166:548e 62. [50] Libert JP, Candas V, Vogt JJ. Modification of sweating responses to thermal transients following heat acclimation. Eur J Appl Physiol 1983;50:235e46. [51] Onaka S, Hori S, Saito N, Shiraki K, Migasena P, Yoshihara H. The role of food habits in physiological adaptations of inhabitants of Southeast Asia to the habitat of tropical countries. Prog Hum Nutr 1978;2:219e35. [52] Lee JB. Heat acclimatization in hot summer for ten weeks suppress the sensitivity of sweating in response to iontophoretically-administered acetylcholine. Korean J Physiol Pharmacol 2008;12(6):349e55. [53] Davis JR, Johnson R, Stepanek J, Fogarty JA. Fundamentals of aerospace medicine. 4th ed. Philadelphia, USA: Lippincott Williams and Wilkins; 2008. p. 209. [54] Nielsen R, Nielsen B. Influence of skin temperature distribution on thermal sensation in a cool environment. Eur J Appl Physiol Occup Physiol 1984;53(3): 225e30. [55] Arens E, Zhang H, Huizenga C. Partial- and whole-body thermal sensation and comfort, part I: uniform environmental conditions. J Therm Biol 2006;31:53e 9. [56] Boutcher SH, Maw GJ, Taylor NA. Forehead skin temperature and thermal sensation during exercise in cool and thermoneutral environments. Aviat Space Environ Med 1995;66(11):1058e62. [57] Oseland NA. Predicted and reported thermal sensation in climate chambers, offices and homes. Energy Build 1995;23:105e15.

R. Maiti / Building and Environment 70 (2013) 306e317 [58] Humphreys M, Nicol J. The validity of ISO-PMV for predicting comfort votes for every-day thermal environments. Energy Build 2002;34(6):667e74. [59] Brager GS, de Dear R. Climate, comfort & natural ventilation: a new adaptive comfort standard for ASHRAE standard 55. In: Proceedings: moving thermal comfort standards into the 21st century. Windsor, UK: Oxford Brookes University; April 2001. [60] Bae C, Chun C. Research on seasonal indoor thermal environment and residents’ control behavior of cooling and heating systems in Korea. Build Environ 2009;44(11):2300e7. [61] Dufour A, Candas V. Ageing and thermal responses during passive heat exposure: sweating and sensory aspects. Eur J Appl Physiol 2007;100(1):19e26. [62] Collins KJ, Exton-Smith AN, Dore C. Urban hypothermia: preferred temperature and thermal perception in old age. Br Med J (Clin Res Ed) 1981;282:175e7.

317

[63] Natsume K, Ogawa T, Sugenoya J, Ohnishi N, Imai K. Preferred ambient temperature for old and young men in summer and winter. Int J Biometeorol 1992;36:1e4. [64] Havenith G. Temperature regulation and technology. Gerontechnology 2001;1(1):41e9. [65] Taylor NA, Allsopp NK, Parkes DG. Preferred room temperature of young vs aged males: the influence of thermal sensation, thermal comfort, and affect. J Gerontol A Biol Sci Med Sci 1995;50(4):M216e21. [66] Indraganti M. Thermal comfort in apartments in India: adaptive use of environmental controls and hindrances. Renew Energy 2011;36(4):1182e9. [67] Singh MK, Mahapatra S, Atreya SK. Adaptive thermal comfort model for different climatic zones of North-East India. Appl Energy 2011;88: 2420e8.